Hello John,
I was worried about “talking behind your back” but it looks like you’re reading this thread too so I’m relieved.
We started discussing this some time ago but I didn’t follow up on that because I didn’t think it was important. But for the purpose of a full argument I will amend that now.
I should maybe clarify that by transaction costs, I do not only mean transaction fees. The latter is merely a subset of the former. The transaction costs include all the costs associated with a transaction that are not the immediate part of the transaction. For example, if you want to pay with cash, you need to carry the corresponding amount of cash with you to the place of transaction. This necessity to carry cash is a transaction cost associated with paying in cash. If you need to spend effort on hiding your money from the government in order to pay with that money in the future, this is also a type of transaction cost.
As I argued elsewhere, the ability of Bitcoin to resist manipulation and confiscation by state is a quantitative, not a qualitative, feature. Of course that’s not to deny it exist, I also laid out several examples myself where this advantage is obvious. My favourite ones are a dead man switch (gold cannot teleport automatically when you’re imprisoned), “cheap horcruxes for everyone” (you can’t copy gold), and split-key signatures (you can’t split gold along a metaphysical boundary).
So, I agree that Bitcoin has an advantage in this respect, and other currencies or gold don’t. But that does not warrant a modification of the regression theorem, neither is a qualitative feature. The purpose of the regression theorem is to explain the formation of price of a medium of exchange (i.e. what Mises calls “objective exchange-value”). How exactly the features of Bitcoin influenced this formation is a matter of empirical analysis, and thus merely loosely connected to economics (Austrian, that is).
I give you that it is possible that without government interference, the decrease of transaction costs might have been insufficient for the price of Bitcoin to emerge (i.e. to counter the network effect of a free market money). It is even more likely that without a preexisting monetary system in general, the price of Bitcoin would not emerge (since there would be no forex markets). But again, this is a comparison of heterogeneous variables and can’t be taken as an economic rule. Nevertheless, from a practical point of view, since we do have a monetary system, and we do have government interference in money, there is no necessity to determine what would have happened in the absence thereof.
One potential mistake in Mises’ arguments is the ignorance of goods which are neither consumption goods, nor production goods, nor media of exchange (as I explained above). Once this is amended, the regression theorem becomes interpretable from a new light.
Selgin calls Bitcoin “quasi commodity money”: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2000118 , however he does not explain the emergence of the price thereof. Murphy, in the video you reference, argues that speculation plus ideology might be a sufficient explanation of what happened, and also says that based on this, it could be argued that Bitcoin is a commodity, albeit an unusual one.
I would simply generalise this into an expectation of the decrease in transaction costs. Bitcoin has a comparative advantage because it eliminates the necessity for money substitutes and provides the same service that historically required a bank clearing system (or newer competitors like Paypal or Western Union). This also includes the features you metion, for example, it’s easy for a government to tell the bank or Paypal to freeze your account. I therefore refer to Bitcoin as “money as service”, because I think that better explains its dynamic features (as opposed to quasi-commodity).
I presented this “money as a service” idea to several economists. Hoppe thinks it’s absurd because
This however is an empirical objection. We do have commoditised services, in particular things that are, like money, subject to the network effect. You can buy online diskspace, for example. I discussed the legal issues (Title Transfer Theory of Contract) with Kinsella to make sure there are no fundamental problems with sales of services (there are workarounds, so it’s not a big issue). I tried to explain the “money as a service” to Murphy too. He is also skeptical like Hoppe but at least I think he got my point better than Hoppe. In Murphy’s lecture we didn’t come to the regression theorem yet, so I’ll have another opportunity to ask him about this.
Once a price (of Bitcoin) has formed, the process is sustainable, because there are people who want to speculate and will continue as long as they expect to be profitable. Volatility or a falling price is not a problem for professional speculators. I for example used to do arbitrage between Mt.Gox and Tradehill during a falling price and was able to be profitable.