In their recently published book, 'The Banker's New Clothes', Anat Admati and Martin Hellwig suggest that the causes of the global financial crisis were similar in some respects to the causes of the nuclear power disaster in Japan in 2011. In the case of the nuclear power disaster, the authors suggest that corrupted politicians and regulators had colluded with the Tokyo Electric Power Company to ignore known safety concerns. They comment:
'When an earthquake and tsunami occurred in 2011, this led to a nuclear disaster that was entirely preventable.
Weak regulation and ineffective enforcement were similarly instrumental in the buildup of risks in the financial system that turned the U.S. housing decline into a financial tsunami'.
It might seem obvious to just about everyone that government regulation of the nuclear power industry is desirable to prevent outcomes such as those experienced in Japan (even though regulation was spectacularly unsuccessful in this instance) but I feel inclined to step back a little to consider why such regulation is desirable. What is the problem that the regulation is intended to remedy in the nuclear power industry?
The obvious answer is that in conducting their business of providing electric power to their customers, there is a risk that nuclear power firms may accidentally cause harm to other people. But that is also true of many other business activities. Firms have an incentive to take precautions to avoid such incidental harm because they know that potential victims can sue for compensation.
So, why is additional government regulation needed in the nuclear power industry? Leaving aside the possibility of nuclear material getting in to the wrong hands, a need for additional regulation may arise because of the potential magnitude of the harm that might occur as a result of a nuclear accident. The harmful consequences of a nuclear catastrophe might be so great that the responsible firm would be unable to pay full compensation. That would pose a problem for government of whether to step in and help the victims, but it also poses the problem of how to ensure that the managers of the firm have a greater incentive to take precautions to avoid a catastrophe that would bankrupt the firm twice over, than to avoid a catastrophe that would bankrupt the firm only once. So, there might be a case for the government to step in to attempt to ensure that adequate precautions are taken.
Is there a similar case for regulation of major financial institutions? When I looked at this question a few weeks ago I suggested that when the failure of one bank leads to loss of confidence in other banks that have taken similar risks might just reflect a process in which the market is taking appropriate account of new information. For example, if a financial institution becomes insolvent because a decline in property values causes a decline in the asset backed securities in its balance sheet, that information could be expected to bring about a re-assessment of the value of assets of other financial institutions. It should not be surprising that those financial institutions that are considered to be at greater risk of becoming insolvent would suffer from a loss of confidence and have greater difficulty in conducting their business. That is the way an efficient market could be expected to weed out firms that can no longer be trusted to pay their bills. There does not seem to be anything in that scenario that is analogous to the harmful pollution released as a result of a nuclear accident.
Why do the authors argue that major financial institutions ought not be allowed to fail? The main reason they give is contagion, which adversely affects the broader economy. When a major financial institution collapses it is unable to meet its obligations to other institutions, which are also weakened. As more financial institutions anticipate liquidity problems and attempt to sell assets, there is likely to be a further decline in asset values. As financial institutions cut back lending, the broader economy is adversely affected.
Those effects on the broader economy would be dampened, in my view, if central banks were doing a good job of maintaining public expectations of steady growth of aggregate demand. Central banks were slow to use tools such as quantitative easing to do this during the global financial crisis. Even if central banks had made a more determined effort to manage expectations, however, it is doubtful whether they would have been entirely successful in countering fears that failure of several major financial institutions was likely to have severe adverse impacts on aggregate output and employment.
The authors make the point that it would be extremely difficult to allow large complex financial institutions to fail without major disruption when they became insolvent. Proposals that they could be taken over by public authorities until they were placed under new ownership would be difficult to implement because these firms have thousands of subsidiaries and other related entities spread over different countries. Separate resolution procedures would be required for different subsidiaries in different countries. Massive problems of coordination would be involved.
Governments seem to have managed somehow to get us into a vicious cycle where fears of contagion have led them to encourage major financial institutions in the believe that they were too big to fail, while the belief that governments would bail them out has led major financial institutions to take excessive risks. If we can't let big financial institutions fail when they become insolvent, perhaps the next best option is to find the least cost way of regulating them to make it less likely that they will become insolvent. That does present governments with problems that are similar to those involved in regulating the nuclear power industry.
In a later post I will discuss Anat Admati and Martin Hellwig's views of how governments can reduce the risk of insolvency in financial institutions that are too big to be allowed to fail.
I am writing this postscript before I have posted the article because I have had some further thoughts about market failure, a concept that I was tempted to mention above. An earlier post about financial crises led to a discussion with Jim Belshaw about the meaning of market failure. During the course of that discussion I conceded that the concept of market failure is of limited use and made the point (attributed to Harold Demsetz) that the relevant choice is not between an existing imperfect market and an ideal norm of a perfect market, but between real world outcomes under current institutional arrangements and a proposed alternative set of institutional arrangements. My new point (new to me anyhow) is that if some feasible outcome is superior to that which exists at present, then past failure to implement the changes necessary to achieve that outcome should be viewed as government failure rather than market failure.