The economics of language: David Hume & valuing Facebook

David Hume is the first great thinker to identify language, law and money as ‘spontaneous’ institutions of social organisation. Hume was on to something quite profound, which remains under-appreciated.[1]

Language, law and money have very similar economic properties. Specifically, the resilience and propagation of these institutions does not reside in some intrinsic, physical value, nor in a promise, nor in the value each individual derives from them. It resides in a network externality. Hume does not use the language of ‘network externalities’. Instead he talks of ‘convention’ and a recognition of mutual interest. His description of the evolution of law remains profoundly insightful, particularly the observation that – like money and language – its value does not reside in a promise (an error which pervades much contemporary analysis):[2]

“I observe, that it will be for my interest to leave another in the possession of his goods, provided he will act in the same manner with regard to me. He is sensible of a like interest in the regulation of his conduct. When this common sense of interest is mutually expressed, and is known to both, it produces a suitable resolution and behaviour. And this may properly enough be called a convention or agreement betwixt us, though without the interposition of a promise; since the actions of each of us have a reference to those of the other, and are performed upon the supposition, that something is to be performed on the other part. Two men, who pull the oars of a boat, do it by an agreement or convention, though they have never given promises to each other. Nor is the rule concerning the stability of possession the less derived from human conventions, that it arises gradually, and acquires force by a slow progression, and by our repeated experience of the inconveniences of transgressing it. On the contrary, this experience assures us still more, that the sense of interest has become common to all our fellows, and gives us a confidence of the future regularity of their conduct: And it is only on the expectation of this, that our moderation and abstinence are founded. In like manner are languages gradually established by human conventions without any promise. In like manner do gold and silver become the common measures of exchange, and are esteemed sufficient payment for what is of a hundred times their value. [Italics added]”

David Hume, A Treatise on Human Nature, Book III, part 2: Of the origin of Justice and Property.

Let’s consider language as perhaps the clearest case in point.

We value language above all. If someone had a patent on language and could impose subscription fees for its use, we would be willing to pay a great deal for the service. Of course, individuals can invent their own language – it costs little to do so, it has no intrinsic value and minimal costs of production. But a personal language is of limited use. The value of language lies precisely in the fact that it is used by others – and it’s value to an individual increases with the number of others who use it. That is precisely what defines a network externality, and also explains why network effects tend towards monopolies.

Laws operate in a similar way. There is no point an individual setting some personal, idiosyncratic rules, which he or she alone follows. Subject to laws being just and efficient, the benefits to any individual of the system of laws increases the more widespread is adherence.

Now, it is of course the case that many of us derive value from the pleasure of using words – epitomised by literature and poetry. These can be private pleasures, but the value of language is not ‘intrinsic’ – words do not have physical properties which can be bought and sold, or put to alternate uses. It is also the case that the state plays an important role in establishing a national language, through standardisation and the educational system. No doubt there is an enthusiastic chartalist out there who claims that the only reason we use the national language is because it is used on tax returns – and were the state to reject its use in official documentation we would promptly abandon it![3]

What about money? The economic properties of money may in fact be closest to those of language, and this may explain the confusion in economics over the fact that something with no ‘intrinsic value’ can be so valuable. Psychological discomfort with this attribute explains both recurrent desires to return to a gold standard, and the more fashionable idea – that money is a debt.

Money in fact works like language. I could create my own, but because no one else uses it, it will have no value. Governments are clearly in a great position to establish dominant networks. They can command widespread acceptance of their money by making it ‘legal tender’, and by requiring it be used to pay taxes. Although these are the means of establishing a network – it is the network externality that is the source of the enduring and resilient value of money (frequently in the face of considerable mismanagement by central banks).[4]

All the artifices which have been used historically to give money ‘value’, are really means to establish networks. As Hume points out, even where money has ‘intrinsic value’ (which really means an alternate value), in the case of gold or silver coins, their value as money soon exceeds their value as gold or silver. The key point is that money is extremely useful, and it’s usefulness lies in its acceptance by a vast number of cooperating, voluntary, users. It’s value to me, like language, resides precisely in its value to others. The process Hume describes is reciprocal, of mutual interest, and increases with the number of users. Disrupting such a network is profoundly difficult, which explains the extraordinary resistance of monetary regimes to repeated abuse (including, for example, hyperinflation). This is the true obstacle to innovations like Bitcoin – if the value of a money resides in the number of users – how can even a superior technology establish value? (Which is not to say that Bitcoin is a superior technology – it may well be inferior.)

Facebook & Visa

The market capitalisation of two companies – Facebook and Visa – listed on the New York stock exchange trade at market values vastly exceeding their book values (ie the value investors are paying for a share in these companies is a high multiple of the value attributed to them in conventional accounting).

To Hume, it would be obvious why: Facebook is ‘language’ and Visa is ‘money’. Hume was more prescient than he could have imagined, because he also identified the source of value of some of the most successful technology companies we have ever seen.

Social media has much in common with language – it uses words, and it uses images, often symbolically. Snapchat is an image-based symbolic language, as is Instagram (the latter owned by Facebook). But the value of Facebook resides primarily in a network externality. It is a source of its monopoly power. Facebook is more useful than a technologically-superior alternative with fewer users. That’s why network externalities are hard to break.

Visa operates like the money it intermediates: it has value because millions of individuals, merchants, businesses, governments etc., all accept Visa cards as payment. This will be the challenge any new payment innovator faces – superior technology alone cannot break a network externality.

So the market capitalisation of Facebook and Visa should not surprise us. One has patented language, the other has a license to print money.

[1] The Austrian economist Friedrich Hayek also attributes this insight to Mandeville, and later to fellow-Austrian, Carl Menger.

[2]Although, astonishingly, given that he was writing in the 1700s, he does indeed discuss “external” goods in this context.

[3]There is probably an economist who believes that language is a liability of the state, because the government has a target for national literacy levels.

[4]It also explains why, in the face of despotism or gross mismanagement, populations will revert to using the currency issued by other jurisdictions, such as US dollars.

About The Author

Eric Lonergan is a macro hedge fund manager, economist, and writer. His most recent book is Supercharge Me, co-authored with Corinne Sawers. He is also author of the international bestseller, Angrynomics, co-written with Mark Blyth, and published by Agenda. It was listed on the Financial Times must reads for Summer 2020. Prior to Angrynomics, he has written Money (2nd ed) published by Routledge. He has written for Foreign AffairsThe Financial Times, and The Economist. He also advises governments and policymakers. He first advocated expanding the tools of central banks to including cash transfers to households in the Financial Times in 2002. In December 2008, he advocated the policy as the most efficient way out of recession post-financial crisis, contributing to a growing debate over the need for ‘helicopter money’.

14 Responses

  1. Nick Rowe

    “What about money? The economic properties of money may in fact be closest to those of language, and this may explain the confusion in economics over the fact that something with no ‘intrinsic value’ can be so valuable. Psychological discomfort with this attribute explains both recurrent desires to return to a gold standard, and the more fashionable idea – that money is a debt.”

    Yep. Good post. But *some* of us economists are already comfortable with the idea that money has network externalities, like language. (Mostly we trace it back to Menger, rather than Hume, but you are right about Hume.)

    The example I like to use: suppose shells are used as jewelry, with a standard demand and supply curve and positive price. Now people start using them as money too. The monetary demand adds to the jewelry demand, making them more valuable. Them fashions change, people stop wearing shells as jewelry, but keep them in their pockets instead. The jewelry demand disappears, but the monetary demand remains, so shells are still valuable, though intrinsically worthless.

    Shells were in fact historically used as money, in many places.

    • Eric Lonergan

      Thanks for the comment, Nick. I couldn’t agree more. Hume makes a similar point regarding gold coins. I also think this raises some interesting questions about government debt – if we start using government bonds as collateral, or indeed as the ‘safest security’, they are no longer debt instruments alone, but have acquired money-like – what does this do to their value, and are there (positive) fiscal implications?

      As an aside, we should point out that once any good or service has additional uses it’s value rises, so an increase in the value of the shells does not by itself establish that money’s value resides in a network externality.

  2. Wim Nusselder

    Language, law and money are not the only institutions with a network effect on their value.
    (I’d suggest to distinguish the institution of property from that of law.
    Property, considering persons and things as linked, and law, rules to abide by, can exist separately.)
    You can even say that the network value of an institution determines how basic it is for society and institutionalisation itself can be defined as the acquiring of netwerk value of a practice or system of practices.
    The most basic institutions of human society could be marriage, the taboo on incest, property and -of course- language.
    The need for and network value of money (and law) comes only with larger scale and complexity of human societies, that complicate keeping track of reciprocity (and custom).

    It remains to be seen how basic money is for modern, globalized society.
    The Uniform Economic Transaction Protocol (UETP) may outdate it, sooner or later, and facilitate even ‘deeper’ globalization and stronger global ‘society’ by doing so.

  3. Peter Sims

    I would be curious to know what your thoughts are about what anchors the value of one unit of currency to a particular exchange value? (Why can we buy a can of coke for roughly a dollar, rather than a million dollars, or a millionth of a dollar? How do we come to agree about that?) Language and law, the other analogous phenomena, do not face this problem, because they do not need to be traded in quantities. If network externalities are the true source of value of money, how we collectively end up deciding what that value is in the first place? Or does this step have to free-ride on some other anchor (clears tax liabilities, is comprised of a fixed amount of shiny metal, can be exchanged at the company store for goods, etc) in order to take on a particular value?

    • Eric Lonergan

      Interesting question. I need to think more about this, but my immediate reaction is a that you are getting at the establishment unit of account function – I have focused on ‘medium of payment/exchange’. Once a price level is settled upon there is a clear convention and it is quite resilient, although institutionalised indexation can link price changes very directly with changes in base money (for example under hyperinflation). It is an interesting question as to how the *original* unit of account is agreed upon. I don’t think it can be taxes – money has to have value already before there is any point in taxing it! I suspect your reference to shiny metal is apposite. The numerical value of the unit of account does get established by reference to a quantity of a commodity or another unit of account. Even if it is *not* redeemable. That is certainly how new currencies are introduced (for example the rentenmark, or indeed the Euro).

      • Peter Sims

        Thanks for the reply! A couple points in response

        1) How do you have a widespread medium of exchange or payment without it being a unit of account? If I can pay a bill with money, rather than barter it for other goods, that means someone has to write and denominate that bill, if only by posting prices. Once that’s done, we have a unit of account, however trivial. Its use as a medium of exchange, and thus all the network externalities, seem to rely on this property as well, and thus it must be accounted for.

        2) I don’t believe it is correct to say that money must already have value for a sovereign to levy taxes in it; this is core to how I understand the MMT claim, and also the fiscal theory of the price level. If a sovereign (credibly) threatens to throw people in jail unless they return some number of fiat tokens at the end of the year, then people will value those tokens because it keeps them out of jail. Those tokens don’t have to exist yet, or be in circulation – the threat of future punishment gives them value. (I think the value of the tokens will then be related to how many of them are needed to keep you out of jail, and some Beckerian calculation about the probability and severity of punishment for nonpayment.) But if the sovereign can credibly commit to levy a tax denominated in those tokens, then they should be able to print and spend said tokens, as seigniorage. This is just a way of making coercion more efficient, but has all the benefits of creating a potential means of payment and unit of account. This is not to provide a historical account, though some MMTers insist it must be, but rather a coherent story about how a currency could come into existence and have commonly accepted value without being anchored to anything but taxation.

      • Eric Lonergan

        Why would the government want the tokens if they had no value? The only point of taxing to raise funds is if the funds have value.

  4. Peter Sims

    The government would not want the tokens per se. They want the credibility that comes with accepting their own (otherwise valueless) token, precisely in order to gain from the kind of network externalities you describe. Seigniorage is valuable: sovereigns with their own token currencies do not have to point the swords at the merchants in the present moment, whenever they want goods, nor do they have to borrow money prior to spending. They can simply print the money in the present, and do the coercion whenever it is convenient. People will accept the tokens so long as they believe this will clear their tax obligations. Accepting tokens to clear taxes is just establishing credibility in a repeated game, trusting in backwards induction from the threat of jail to give value to otherwise worthless pieces of paper. (Or broken sticks, or stamped pieces of base metal, or numbers on a spreadsheet.)

    • Eric Lonergan

      ‘Clearing tax obligations’ is a semantic conflation – it just means ‘pay’.

      A useful medium of exchange is also a useful unit of account. I don’t think anyone would disagree that. Taxation does not establish value as a unit of account nor value in exchange. It can help establish a dominant ‘money’ – although a legal system with effective legal authorities would do the job too! But the value of money resides in a network externality which makes it qualitatively different to ‘debt’. The value of a debt resides in the ability to do the borrower to meet interest and capital repayments.

      • Peter Sims

        “Taxation does not establish value as a unit of account nor value in exchange.” But taxation in sovereign currency does both. If a sovereign levies taxes, they are explicitly giving people a choice: Pay a particular amount of money, or face the penalties. Those penalties will have associated disutility, which people will value in some ratio compared to their time and goods. (I might work for a month a year to pay my taxes, but not for eleven months; at that point, I’d rather just go to jail.) Everyone who needs to pay taxes gives up some of their labour and goods to obtain the tokens through trade, and since everyone is making this calculation (and can trade the tokens freely) even an entirely arbitrary token currency should converge on an equilibrium value; a value in exchange. It depends on ordinary supply and demand logic: how many tokens are issued, how many the sovereign demands in taxes, and what they do to you if you don’t pay up. If tokens are plentiful relative to the tax burden (how many you need pay to stay out of jail) then their value is low, and vice versa. This doesn’t have to be the circulating money, but it does have an exchange value created by the punishment for non-payment, and it does have to be a unit of account, at least in terms of accounting for the sovereign’s taxation.

        Nick Rowe’s excellent comment at the beginning involved seashells first being used as jewellery, beginning its life as a commodity currency, then losing that anchor. The “inherent” utility of fashionable seashell jewellery to people serves the same purposes there as the disutility of punishment does here in establishing an exchange value. I suspect such a seashell currency (and its network) would be very unstable once it became unfashionable, as there is no longer any reason except customary trust in the secondary market to give one seashell any particular value. So, too, with a fiat currency no longer anchored to taxes. If network externalities are the key, presumably you would predict otherwise, that its value should be quite durable?

      • Eric Lonergan

        ‘If network externalities are the key, presumably you would predict otherwise, that it’s value should be quite durable?’ Indeed. That is one of the points made in the piece – and it is reflected in durability of fiat money in the face of occasionally appalling abuse by issuers.

        You have raised a very interesting new point, however. Money must have value prior to government raising taxes or the government would have no incentive to tax. It would be interesting to see how the history of taxation evolved. I’m guessing the establishment of money precedes the establishment of taxation, because administrative organisation is later in social evolution than money.

Leave a Reply

Your email address will not be published.

* Checkbox GDPR is required


I agree