Saturday, December 20, 2014

Bitcoin Is Not Like Yap Stone Money

For centuries the people on the island of Yap in the South Pacific used a currency that was both useless and inconvenient, which was probably part of the reason that they didn't even care whether they actually received the currency when somebody paid them in it.

Put like this, Yap money sure sounds weird, yet some economists and bitcoin experts claim that there are deep similarities between this primitive money on the one hand and the most innovative and hi-tech money that we know of, bitcoin, on the other.

To understand why this need not in fact be an unreasonable claim we need to have a closer look at the key features of Yap money that I hinted at above. 

To understand why the claim nonetheless is probably inaccurate it is necessary to show that there is surprisingly little evidence to think that Yap money in fact functioned the way economists think it did.

Yap Stone Money

Stone money
The people on Yap used stones as their main form of currency. The stones were imported from other islands and they came in different sizes and had different values based on a number of different factors, among which size was a key one. The smallest stones could easily be carried around but the biggest stones were huge. Like, really really huge. There are two main features of the stones that made them such remarkable money.

1. The stones had no non-monetary value. The only thing that gave the stones their value was their use as money. While it is often thought that a good can only come to be used as money if there is a prior demand for the good based on a prior non-monetary role (e.g. if commodities such as grain - or cigarettes in prison camps - come to be used to enable indirect exchange) Yap stones seem to be a clear counter-example as they never seemed to have had a non-monetary role in Yap society. 

Nor was there a government that forced people to use the stones to pay taxes, which could have been an alternative way to create demand for a good that seems otherwise useless. So Yap stone was that rare breed of private fiat money.

2. Selling stones for other goods in the economy did not require actually physically exchanging said stones. Normally when you pay for something you physically transfer whatever is used as money or as a claim on the money (coins, bank notes, shells, IOUs). But in the case of Yap some of the stones were so big and unwieldy that physically transferring them with each transaction would make transaction costs very high indeed. 

So the idea is that instead of doing that, the islanders came up with an ingenious solution: If islander A sold a stone to islander B in exchange for some other commodity, the stone could stay right where it was and did not have to be physically transported from A to B. The islanders would simply note the transfer and agree among themselves that the stone now belonged to B instead of A. If B then in turn wants to sell the stone for something else, others know that he is the rightful owner of the stone, whereas if A were to try to do the same thing buyers would not accept it because it was not hers to sell. So the islanders were using a kind of shared and decentralized ledger to keep track of the ownership and changes in the ownership of stones.

The fiat nature of the stones and the use of a shared ledger to prove and transfer ownership of them may have been related to each other: If the stones did not have a non-monetary use, then there would have probably been little appeal in physically possessing them as long as there was an alternative way of proving and transferring ownership, such as through a shared social ledger.

These two features, moreover, will sound familiar to anyone acquainted with bitcoin. After all, bitcoins have no or only a negligible non-monetary value and the way in which (changes in) ownership of bitcoins is kept track of is not by physically transferring bitcoins but by the seller announcing to the world that her bitcoins now belong to the buyer, i.e. that her bitcoin balance is to be reduced by a certain amount while the buyer's balance is to be increased by the same amount. All these proposed transfers are recorded in blocks and the chain of validated blocks functions as the social, decentralized ledger that keeps track of changes in the ownership of all bitcoins in existence.

In the case of Yap, to the extent that the ledger was maintained in a decentralized way – by each islander having a copy of the ledger in their mind or stored otherwise, and by each islander being kept up to date of any updates in the ledger - there were strict social, cognitive and computational limitations on the size of the ledger that the islanders could share. Moreover, they would actually have to agree with each other on the content of the ledger and any changes therein. Bitcoin's blockchain technology on the other hand did away with these limitations and vastly increased the potential size of the shared ledger that could be maintained in a decentralized way. For these reasons bitcoin can be regarded as Yap money on steroids, as Yap money for the digital age.

The problem with the analogy between Yap stone money and bitcoin, however, is that the description of Yap money on which it rests has little basis in reality. Specifically:

1. It is not at all clear that the stones really did not have any non-monetary value.

2. There is little if any reason to think that physical possession and transfer of stones was not required in transactions with some Yap stones and that the islanders in part or in full relied on a shared ledger to keep track of (changes in) ownership of said stones instead.

3. Even if the islanders had used such a shared ledger, Yap stones could still have been used as money without reference to a ledger while the same is not true for bitcoins. 

Did Yap stones really not have a non-monetary role?
JP Koning provides an excellent overview of prominent economists who have claimed that Yap stones were pure fiat money, that they had no non-monetary role in society. Koning also discusses the work of Dror Goldberg who provides ample evidence in his paper Famous Myths of Fiat Money to show that there are actually good reasons to think that these claims are simply false, that the stones in fact did have non-monetary roles – in particular that they had significant aesthetic and religious value - in Yap society.

I should note, however, that even if the stones did have such non-monetary significance this does not mean that the monetary role was based on a prior non-monetary role. It could have been the other way around, i.e. it is also possible that it was the monetary use that made the stones valuable and that it was this value that caused them to be used as ornaments or in religious contexts, rather than the other way around. The stones may have started being used as money simply because they were scarce, impossible to forge or replicate, easy to carry around (the smaller ones anyway). As their use in exchange and hence their value increased people could use them to show off or for gift-giving or tributes, and their growing significance in society may have also given them a place in the mythological, religious practices and narratives of the community.

If this is what actually happened (and we don't really know either way), then the way Yap stone money is typically portrayed would still be incorrect as the stones would have non-monetary value, but at the same time it would be incorrect to think that the non-monetary role was prior to the stones' monetary role. Moreover, if this scenario is accurate, it may also tell us something about the possible future uses of bitcoin, in that if bitcoin were to continue and grow as a money it may also take on non-monetary roles based on the value of the prior monetary role and significnce in society. It is not immediately obvious what a bitcoin necklace would look like though!

Alternatively the stones may have come to be valued neither because they were used as money nor because they had some immediate non-monetary use, but simply because people may have an innate tendency to collect and value and trade such scarce things, even if these things don't have a role outside of being valued as such collectibles. Such use may form the basis of later monetary and/or non-monetary roles. It may also suggest that even if bitcoin does not succeed as a medium of exchange bitcoins could still continue to be valuable, just not as a currency but more as a non-monetary collectible similar to baseball cards, first editions of books etc.

Was there really a Yap ledger?
What is the evidence for the claim that the Yap islanders were content to simply make a mental note of changes in ownership of stones rather than physically transfer them, and thus that the Yap monetary system used something akin to a shared social ledger to keep track of and transfer ownership of the stones? As far as I can tell this portrayal of the Yap monetary system is based on the following two paragraphs in the original account of Yap stone money by Furness:
The noteworthy feature of this stone currency is that it is not necessary for its owner to reduce it to possession. After concluding a bargain which involves the price of a fei too large to be conveniently moved, its new owner is quite content to accept the bare acknowledgement of ownership and without so much as a mark to indicate the  exchange, the coin remains undisturbed on the former owner’s premises.
My faithful old friend, Eatumak, assured me that there was in a village nearby a family whose wealth was unquestioned—acknowledged by everyone—and yet no one, not even the family itself, had ever laid eye or hand on this wealth; it consisted of an enormous fei, whereof the size is known only by tradition; for the past two or three generations it had been and was at that time lying at the bottom of the sea! Many years ago an ancestor of this family, on an expedition after fei, secured this remarkably large and exceedingly valuable stone, which was placed on a raft to be towed homeward. A violent storm arose and the party, to save their lives, were obliged to cut the raft adrift, and the stone sank out of sight. When they reached home, they all testified that the fei was of magnificent proportions and of extraordinary quality, and that it was lost through no fault of the owner. Thereupon it was universally conceded in their simple faith that the mere accident of its loss overboard was too trifling to mention, and that a few hundred feet of water off shore ought not to affect its marketable value, since it was all chipped out in proper form. The purchasing power of that stone remains, therefore, as valid as if it were leaning visibly against the side of the owner’s house, and represents wealth as potentially as the hoarded inactive gold of a miser in the Middle Ages, or as our silver dollars stacked in the Treasury in Washington, which we never see or touch, but trade with on the strength of a printed certificate that they are there.

And that's it. Those two paragraphs are, as far as I can tell, all that Keynes, Friedman and so many others who came after them cite as a basis for their beliefs. Not only is this very little evidence, there are also some serious problems with it:

1. Furness simply gives no examples to back up the claims he makes in the first paragraph. 

2. It seems that the only example he does try to give is the one in the second paragraph. The problem is that he doesn't actually show what he and so many after him think it shows, i.e. that it is an example of what he mentions in the first paragraph. Specifically, he doesn't show that the stone at the bottom of the sea actually continued to be traded. He only says that it has been in the same family for at least 2 or 3 generations and that their wealth is unquestioned, which is pretty much the opposite of its being traded. Also, the person he learns this from may not have been reliable, according to Goldberg.

There are also more theoretical reasons to be skeptical of the way in which the Yap monetary system is portrayed. If the Yap stones were the main currency on the island this means first of all that the economy was relatively well developed in that there was so much trade going on that there was a need for money at all, and it also means that the Yap stones were used in a large share of all the transactions. 

On Twitter Nick Szabo points out though that the community and number of stones were small enough for islanders to mentally keep track of ownership of stones. Moreover, there were violent punishments for lying, which would be a way for the islanders to maintain consensus about the ledger. If so, then the theoretical objection I mentioned above would be unfounded. I'll look into this in more detail.

That in turn means that it quickly becomes a lot of work to keep track of all the transactions and hence changes in ownership of the stones taking place, or even just the ones involving larger stones. How could the islanders possibly keep track of all this in a decentralized (or even in a centralized) way? And even if they somehow managed to do this, would there not have been ample empirical evidence of this?

A Bitcoin without a Ledger
Even if the Yap islanders had in fact used a ledger in the ways described above, there would still be a crucial difference between Yap stone money and bitcoin: Yap stones are scarce, impossible to forge or double spend, and easily identifiable as authentic. These features are all a result of their physical nature and without reference to or dependence on a ledger. Moreover, ownership - or at the very least possession - of the scarcity of Yap stones can be transferred by simply transferring the physical stones.

Bitcoin on the other hand relies entirely on the blockchain for all of these things. It is only because of the blockchain that bitcoins are scarce, costly or impossible to forge or double spend, and that their authenticity can be verified. Moreover, transferring ownership or at the very least possession of bitcoins can only be done on and with reference to the blockchain.

More fundamentally: The ledger and the bitcoin are inseparable: there is no bitcoin without the blockchain and no blockchain without the bitcoin. Unlike in the case of Yap stones in the case of bitcoin the ledger is an inherent part of its ontology. 

While with Yap stones you can use a ledger to keep track of who owns which stone, with bitcoin you need a ledger for bitcoins to be able to exist at all.


In short: Bitcoin is not like Yap stone money.


  1. Koen, thanks for sharing this article. As always I am impressed with your analytical capabilities. But I am left begging for more, and asking myself "So what? Why does it matter"? You begin with a nice summary of arguments that compare Bitcoin to Yap stone money, and then you go on to show how Bitcoin differs from Yap stone money, namely in that the Yap stone ledger is optional (a feature of the system) whereas with Bitcoin the ledger is a core system requirement (part of the MVP, so to speak). So my question: Why is this important to you? Why does this matter? What are the implications of this analysis? -Ryan

  2. Bitcoin hesabı açmak mı istiyorsunuz? Tıklayın:Bitcoin hesap açma