Saturday, April 27, 2013

Fed power limits after bitcoin (part 2)

The basic idea expressed by if some economy somewhere experiences hyperinflation and responds with de facto adoption of bitcoin (or some other similar crypto based currency), this would stabilize the value of that currency and engage network effects that would nearly guarantee that all subsequent currency changes would choose the same crypto currency, eventually making it the dominant strategy and crowding out all national fiat currencies. Reading that post got me thinking.

I think of bitcoin as a virtual commodity. I was curious about what effect it could have on central bank policy effectiveness while it is still marginal. I now think this boils down to the question of what effect mutual fund-style checking accounts have on central bank policy effectiveness, since those sorts of accounts already use non-monetary assets as if they were part of the payment system. So the question is more general and less interesting. The spread of money market mutual fund accounts with checking privileges that happened in the 1980's did not particular rein in the federal reserve.

Bitcoin has two uses currently. People can use bitcoin to make anonymous transactions, with no paper trail and ignoring national boundaries and legal restrictions. Bitcoin transfers may be less expensive than other transactions, even with a paper trail, by reducing or eliminating bank transfer fees. These provide bitcoin with value, but they will disappear if people fear a cash too much. If we ignore issues around hassling with the mechanics of the transfers, the minimum value of bitcoin to a particular person equals the value of the fee that it avoids or the value of avoiding a paper trail. . 

I think it is interesting that bitcoin is so easy to transfer, whether locally or internationally. I guess I am concluding that even in a severe hyperinflation, bitcoin would not be likely to replace dollars. But the d
ollar is a special case because it is the default global reserve currency. Currency boards using the dollar as the reserve currency seem like the current alternative to national fiat currency, and are sometimes used to end or avoid hyperinflation. Under what sorts of circumstances might a government experiencing hyperinflation create a currency board based on bitcoin rather than dollars or euros?

Can a commodity whose value in terms of other commodities is varying wildly become the unit of account? How much variation in value can a unit of account tolerate before it breaks down or is replaced?

My Debt to Robert Anton Wilson

I decided that Wilson's thinking had enough influence on me that I owe him a mention here on the blog. I enjoyed Wilson's sense of humor and imagination, but as an author of fiction he was nothing special. Maybe he wasn't so special as a writer of nonfiction either, but I encountered him and his unusual epistemological attitude at a fortuitous moment in my life. My subsequent exposures to experiences and orthodoxies have all been colored to some degree by Wilson's peculiar brand of gullible skepticism and suspicion toward orthodoxy. There are other authors who can offer a variety of this attitude, and I've probably collected a few. Perhaps I even had a predisposition toward this friendly skepticism/crude Bayesianism before I encountered Wilson, but he made me conscious of the issue before anyone else did. This seems important to me.
I'm not sure whether I would recommend his books to anyone at this point or not. I love him, but he has plenty of baggage. I'm tempted to think that one good article (or blog post) could encapsulate perfectly what Wilson gave me. But on the other hand, if he'd just written a brief article on epistemology, I would never have read it, and even if I had, maybe the rhetorical effect would have been less.
I scratched together some notes and began googling to try to figure out whether Wilson mentioned Bayesianism in his books or articles. My skill with google may not be sufficient to the task, but I can't find any evidence that he did. Plenty of other people mention Wilson and Bayes, though, including this reasonably pleasant article on Bayesian probability: Wilson is mentioned briefly in one of the comments.
Another blog post criticizes Wilson as possessing/transmitting knowledge *about* mysticism rather than possessing mystical knowledge, and claiming that he was a "dabbler". It's true that Wilson was eclectic, an empirical and practical tinkerer interested in field stripping the human mind and seeing if he could hack it. Some of his nonfiction included exercises he seemed to think would give results (I was never patient or empirical enough to actually try them), though maybe this isn't the sort of mystical knowledge that the poster intended. Wilson considered mysticism to be a primitive form of psychology, so from his standpoint there is no problem.
I do wish I knew which of his works would be most appropriate as a "gentle introduction" to Wilson. "Illuminatus!" I think was his most well-known book, but having reread book 1 fairly recently, I don't think it fits the bill. (Though that was my introduction, that book by itself did not have much impact on me, as I just concentrated on the jokes my first time through. The second time I could see some of the philosophy peeking through, but it was a bit too subtle for this purpose.) I spent some time dredging through YouTube videos, but I think they mostly would make him seem a bit nutty to someone not familiar with his sense of humor. So it's a problem, I have not solved it. Suggestions in the comments?

Tuesday, April 23, 2013

Fed power limits after bitcoin (part one?)

This post was inspired by a post on "In Crypto We Trust" . Many supporters of bitcoin hope that it will someday completely replace conventional currency, and this obviously would end the economic power of central banks such as the Federal Reserve. In this post I want to look at the obstacles along that path, and what direction it might follow in the short term. Particularly, I am curious whether wide-spread use of bitcoin for transactions would inhibit the power of the central banks.
Bitcoin is an online peer-to-peer currency and payments system. It is not associated with any government, and due to its p2p architecture, no single government can shut it down.
Traditional econ describes money primarily in terms of 3 functions: medium of exchange, store of value, and unit of account. Despite various factors opposing innovation, the monetary system has been moving toward separating these functions for decades. This trend may accelerate as a result of the arrival of bitcoin. It is already difficult to say what is and isn't money, and will only become more difficult.
The rise in popularity of money market mutual fund accounts in the early 1980's illustrates the idea that we can use nearly any asset as a store of value and medium of exchange. People using these accounts can write checks against them like an ordinary checking account. But the interest earned comes from the notes and short term bonds owned by the fund, rather than an arbitrary deal with a banker. The mutual fund disintermediated the banks, allowing the share owners to increase their risks and rewards. The checks can clear in various ways. The most general approach is that it would go through the bank clearing house, like an ordinary check. But it is also possible that the MMMF could transfer securities or other assets to another fund to clear their balances. (In the case where the clearinghouse uses dollar denominated bank deposits or notes to clear checks, dollars act as the medium of exchange, but in some cases stocks or bonds may be exchanged instead, at least in principle. I'm not sure whether this happens in actual practice. I suppose it would in the rare case that I write a check to someone who then deposits it in the very same MMMF that it is drawn from, so that the check could be cleared merely by updating the balances of both our accounts.) This practice (having checks or debit cards that draw on the balance of a mutual fund rather than a conventional bank account) quickly spread to ordinary stock mutual funds and others, so that we can now use almost any financial asset as a store of value and medium of exchange. So, are all these assets "money"?
We have not discussed money's function as unit of account. This means that when I do my double entry bookkeeping, my credits, debits, assets and liabilities will all be recorded in terms of money values. (For the uninitiated, double entry bookkeeping was an essential business innovation that still provides the foundation of accounting and lets entrepreneurs know when they are making profits or taking losses. The unit of account allows all the disparate elements of a business to be combined into one number of great significance. If double entry bookkeeping stopped working, society would crumble. Fortunately, most of society can ignore this, since it is not going to stop working.) And in practically every case, that means an item's sticker price will be denominated in the same unit. This allows us to take "money" out of our pockets and pay for something in a store. So most businesses in the U.S. use dollars, companies in Europe tend to use Euros, etc. This, currently, is the primary difference between the money in your bank account and the shares in your mutual fund - one has a fixed price in terms of the unit of account, the other has a market price that changes from moment to moment.
We can imagine a world where dollars are used as the unit of account but dollars are not used as the exchange medium/store of value by thinking about a world identical to ours except that all payments were made by check or debit card, and that all checks and cards were associated with MMMFs or stock mutual funds, etc., where value is stored as stocks or bonds or whatever. These concepts came bundled together in traditional money, but they could conceptually be separated, and as payment systems have advanced, the connection has become looser.
From this standpoint, bitcoin doesn't look like money, but nothing looks like money. From here, bitcoin looks like a digital commodity. It's big advantage is that it can be transported by wire, radio, or any other medium of communication.
What would need to happen for bitcoin to become the unit of account for some business? It would have to be true that this business did enough of their transactions in bitcoin that this approach would be easier. For a retail business, that would mean most of their customers would walk in ready to spend bitcoin, and most of their wholesale suppliers would be willing to accept bitcoin. Also, the unit of account becomes less useful as its value (in terms of commodities) becomes less predictable. Double entry bookkeeping doesn't work well if the value of a business's inventory varies wildly in terms of the unit of account. Hyperinflation requires store owners to frequently update their books and alter their sticker prices (or stop putting prices on goods, which also reduces efficiency). And that is simpler than the case where the value is varying wildly in both directions. So, for bitcoin to become popular as a unit of account, FRNs would need to experience strong inflation or outright collapse, and the average value of commodities in terms of bitcoin would need to experience a period of stability. This will be difficult, if not quite impossible. First steps would include a) bitcoin becomes widely acceptable and widely used, and b) bitcoin's value stabilizes in terms of conventional currency and other commodities c) conventional currency collapses, while bitcoin stays stable with respect to commodities generally. Note that if the collapse of the conventional currency boosted the commodity value of bitcoin, this might actually delay or even prevent businesses from flipping over to using bitcoin as their unit of account.
Can bitcoin become widely acceptable without displacing traditional currencies as the unit of account? Transfers of bitcoin are potentially very convenient. I think what is needed is an application that allows the parties to a transaction to easily determine an agreeable exchange rate between bitcoin and conventional currency. This seems doable, though it is not clear that either conventional businesses or their customers will see much advantage to using it.
In this post I've laid some groundwork but not yet found the answers I am looking for. Rather than delay posting, I want to post this now and I hope that I can finish the job in a future post. Meanwhile, please comment if you have hints, questions, or suggestions. To reiterate, the question I want to answer soon is: If somehow bitcoin became a widely used medium of exchange in the US, but the dollar remained the unit of account, would that limit the central bank's powers in any way?

Monday, April 22, 2013

Lessig on Corruption

I almost agree with Laurence Lessig's recent TED talk, "We the People and the Republic We Must Reclaim." I like his diagnosis, though his argument has some weaknesses, and his prescription fails.
A short version of his idea is that funding for political campaigns distorts the incentives of elected officials to the detriment of serious reformers of every type. He did not give details for specific reforms he thought would solve this, though he mentioned the fair elections now act, the American anticorruption act, and the Grant and Franklin project (from Lessig's book). He also mentioned existing reforms in the states of Arizona, Connecticut, and Maine, though he did not give details.
I wish I could support Lessig's diagnosis whole-heartedly, but I have some quibbles. He accuses funders of acting self-interestedly, and implicitly his analysis depends on the attitudes and beliefs of funders differing systematically from the public in general. His first accusation seems rhetorically sound, since it appeals to the common sentiment against greedy wealthy people. But unfortunately, political scientists have done empirical research that shows that people do not generally follow strict self-interest when voting. We can generalize voting behavior better by saying that people generally identify with some group and vote as a block. Unless we convince ourselves that rich people vote as a single block and adopt policies that are at odds with general opinion, it's not clear that funding has the effect Lessig thinks it does. In other words, maybe George Soros balances out the Koch brothers, and "fixing" funding won't fix policy.
According to Bryan Caplan, corruption happens "on the margin" of policy. I think that means that, if legislators face a range of policy alternatives, some that would harm special interests in favor of public interests, some that ignore public interests for the sake of special interests, and others that favor special interests while not obviously harming the public, lobbying will favor the last category. While this is less than ideal, it means that grossly unpopular policies will not be chosen often. If we believe Caplan, bad policies have more to do with popular biases than with corrupt elections. In effect, Caplan's claims make funding reform a side issue, not the central issue that Lessig thinks it is.
I was disappointed that Lessig spent so much time trying to convince us there is a problem, and so little convincing us that a solution exists. I think I'd have found his talk much more convincing if he'd spent more time describing the proposals he mentioned, and especially discussing the effects of the existing examples of state level reforms. I'm an empiricist, so there's not much I like better than a working model.
Lessig tried to "think outside the box" but didn't go far outside. The federal government is 18th century technology. Lessig wants to patch a bug. I think we might want to consider an upgrade. Government is the phone company before deregulation. Even if we were able to get the incentives right, information problems and coordination problems remain. The 20th century mania for mass production motivated the centralization of power and control. I hope that the 21st century will embrace the Internet and a new federalism, returning to the discarded idea that states should be "laboratories of democracy" instead of "one size fits all." The U.S. can't have two foreign policies or two military budgets, but many policies that have been "federalized" need not stay that way. If we really cared about "getting it right" we would allow more experimentation.
Instead of giving voters vouchers to increase their influence over primaries and advertising, why not reduce the power of legislators? NH has 400 legislators, who can afford to bribe them all?
But none of these reforms is likely to be considered. Politics is frustrating because if you lose you have mostly wasted your time. I need to figure out a way to change the world that actually works, where effort gets paid by results.