The Australian industrial cycle, and the consequent financial panic, that occurred in 1893 is widely considered to be the most condemning empirical case study for free banking. A highly cited paper is Charles R. Hickson and John D. Turner, “Free Banking Gone Awry: the Australia Banking Crisis of 1893,” Financial History Review 9, 2 (2002), pp. 147–167. It is a well-researched piece that reviews the literature and finds that Australia’s relatively deregulated banking system (far more deregulated than most, if not all, other contemporaneous banking systems) had acted frivolously and, therefore, was unable to cope with the sudden depreciation in the value of its assets in and around 1893.
Australia’s crisis of 1893 deserves much more research than I afford it here — and I hope to be able to conduct that research in the not-too-distance future —, but I actually do not find the Hickson and Turner paper all that damning. In fact, there is some evidence presented in the paper that I think can be interpreted from a slightly different angle. Take this post as only food for thought; it is certainly not meant to be conclusive, in the sense that even I am not sure on the validity of the points I raise here. This is really a stepping stone to further investigation.
Some points to start out with,
1. Free banking and Austrian intertemporal discoordination: Around a week and a half ago, I posted an explanation of how the theory of free banking does not conflict with the Austrian theory of industrial cycles (what I like to call “severe intertemporal discoordination”). Some commentators, without actually engaging the brunt of the theory I presented, argued that the 1893 crisis is evidence enough that free banking produces business cycles. This, actually, is not so clear, and even Hickson and Turner accept the evidential ambiguity. Rather, they prefer to take off from the fact that the Australian banking system did not respond well to what could have been an “exogenous” shock (exogenous to the banking system, at least — many blame capital inflows from Britain [I am not so bought, but it is something I plan to investigate]). Further evidence is in the restrictions on note issue placed on Australian banks at the time (“bank note issue was limited to paid-up capital” [p. 153]), thus an over-issue of fiduciary media does not seem a likely cause.
2. Response as Evidence of Weakness: As aforementioned, Hickson and Turner suggest that while the fluctuation may have been exogenous, the real test for free banking is in the response to these types of problems. Given what occurred in Australia, it is not clear that a free banking system can respond well. Even assuming that banks were not hindered by certain restrictions or government actions, I am not so sure that this is a fair basis for assessment. If banks have been led to make unhealthy investment choices (I will return to this below) and find themselves liquidity strapped when the true nature of their investments are revealed, it is not fair to attack free banking on the notion that a banking system ought to be able to respond to any crisis. Assume for a minute that a free banking system is more optimal than a regulated system (humor me). Hickson and Turner’s logic can lead one to conclude that a regulated system is preferable if only because it can respond to these kind of exogenous, “artificial,” shocks. But, it would make more sense to opt for a free banking system and deal with the actual causes of these exogenous shocks.
3. Risk: It is claimed that Australian banks made risky investment decisions, especially in real estate. What is not clear, though, is whether the true risk was assessed ex ante or ex post. This is important, because if the risk was revealed ex post then it could mean that banks were not taking these risks knowingly. It is fairly similar to the most recent recession in the United States, where most assets had been rated AAA by the “big three” Federally recognized rating agencies. It is not exactly the same, but the risk in lending to those investing in what turned out to be volatile markets may have been concealed. For instance, we know that the prices of many of these assets were artificially maintained by further capital inflows from Britain. Banks did not have any information concerning their true value, especially in the case of an unexpected flight of British capital (which occurred in the first few years of the 1890s, due especially to financial troubles in Britain [and the United States]). Most of the banks which failed had large British deposits. (While I have not yet looked at the data closely, I am aware of some of Selgin’s comments regarding the “risky” behavior of Australian banks — if Selgin is right, it is further evidence that the nature of the risk was not ex ante recognized.) More evidence is found in the fact that banks had invested by making advances to Australian construction companies (p. 159), which may be construed as evidence of the fact that there were no expectations of a future crash in prices. The signals at the time may just have been deceiving, and one can hardly blame the banks for that.
4. Restrictions on Response: One of the most powerful tools unregulated banks dispose of in response to information runs is the ability to issue notes. While I am not exactly sure on the limits to Australian note issue, it is clear that some restrictions were in place. Given that banks could only issue notes against paid-up capital it is likely to be the case that they simply could not issue notes in response to demands to withdraw deposits of inside money. Very often, this is enough to inspire a banking panic (a reason why many academics support deposit insurance). Another (much weaker) argument can be made with regards to the decision to give certain banks the privilege of issuing what was suddenly deemed to be legal tender. These banks were able to protect themselves by issuing currency, but it may have also meant that there was less interest in inter-bank lending; in other words, it may have undermined any incentives. I call this argument weak because: (a) most banks were not interested in having their currency considered legal tender and (b) most banks were affected, making inter-bank lending less flexible of an option (which is basically what Merrett  argues [Hickson and Turner, p. 162]).
5. Intervention, Survival, and Deposit Issue Restrictions: This is a comment on two different things. First, Hickson and Turner argue (correctly) that some intervening policies in New South Wales managed to save a good number of banks. But, this is not enough to convincingly argue that this is preferable to free banking. Artificially supporting unhealthy assets represents a social cost. The failure of these assets should not be assumed to be necessarily bad. Second, the authors note that countries which restricted issues of funds provided through account deposits did not suffer similar banking crises. This ought to be intuitive. If bank cannot invest at all, there is no investment that can fail. But, again, this is not enough to argue that this is a preferable policy. In fact, maximizing healthy investment is preferable; if there are exogenous (again, to the banking system) factors which cause fluctuations in output these should be solved.
Like I said, all of this should be taken only as food for thought.
Also, does anybody have electronic copies of Merrett (1989) and Merrett (1993) available? I am having trouble finding them online. Thank you!