# Regression Theorem and Bitcoin

I’ve been thinking through this for a few days and wondered if you had any insight — is the Regression Theorem a praxeological statement, or is it a heuristic device?

If it is a praxeological statement, then it must be true, and we can deduce *a priori* truths from it on that basis (e.g., IF something arises on the market as a medium of exchange, THEN it had value on the market prior to becoming a medium of exchange). This seems to be how many Austrians have treated regression, but it doesn’t seem clear that this is the intent of the discussion in *Theory of Money and Credit*.

If, on the other hand, it is a heuristic device for description and analysis, it can still be useful as an insight into the origins of money, but *a priori* truths cannot be deduced out of it. I’ve looked, but have found very little discussion of this in the Austrian literature. Do you have any thoughts on this? Have I missed something major?

Here was my response:

In my view, the regression theorem is apodictic, praxeological. This brings up the question of the bitcoin. It is not yet money. It is not now a generally accepted means of final payment. But it is now at least a quasi money. More than just a few people treat it as a money.  Probably, the govt will soon blow this out of the water with regulations, taxes. But, if not, it might become a money. If so, would this be a violation of the regression postulate? Yes, if we interpret it as saying that nothing cannot become a money unless it was at one time a valuable COMMODITY. Of course, bitcoins were never a valuable commodity. But, if we more sympathetically interpret the regression theorem not in terms of a commodity, but in terms of SOMETHING of value, then when and if bitcoin becomes a money, it will not contradict the regression theorem for, surely, before it became a money (if it does) it was SOMETHING of value, albeit not a commodity, because it cannot be denied that some people valued it.

I’m not aware of any formal discussion of this in the literature. But, I went to the Mises web, and found this: