I have recently read a report on the banking crisis (“gambling with other people’s money” henceforth referred to as GWOPM) written by Russ Roberts of George Mason University. The report clearly was not aimed at the academic community. However I still found the report lacking in precision of formulation of arguments. I also found that there was not enough evidence provided to back up the arguments made.
I am sympathetic to Russ Roberts’ motivation. Russ Roberts is a classical liberal and I share many political views with him. I think Russ and I would agree that in order to promote classical liberalism we must have good arguments for it. I think that this entails that any argument we make for classical liberalism must pass the most stringent tests available. Russ Roberts has a PhD and is a professor of economics so some may say that I am out of my league to argument with him. But this is not about a conflict between me and Russ Roberts, but rather a conflict of ideas (ideas of how to best express an understanding of the banking crisis). Because this is a battle of ideas it makes no difference who is presenting them. Ideas fight in an intellectual dimension and we are all just observers.
I think that GWOPM has many flaws. I think that overall GWOPM thus does not serve the classical liberal cause very well. Firstly I will outline on what basis I made this judgement. Then I will show some evidence to support my judgement.
Any theory is either disproved or reinforced with close scrutiny. Scrutinising of a theory is done in two ways.
Firstly the theory must be checked to be logically valid and verifiable. A theory is valid if the theory’s conclusion logically follows from the theory’s premises. A theory is verifiability only if it can be falsified. This means that the premises or the conclusion of the theory is measurable or quantifiable. This implies that theory must be formulated in a very precise way so as to be able to quantify varies effects and their causes. Otherwise it would be impossible to judge its logical validity or its verifiability.
The second way a theory can be scrutinised is to actually measure the premises and the conclusion in the real world. A theory can be logically valid and verifiable but disproved by real life data.
A theory can never be positively said to have been proven by real life observation (see David Hume’s problem of induction and Karl Popper’s Conjectures and Refutations). It may be only falsified or not falsified. However once a theory has withstood many tests against real world observations it is deemed to be the closest to truth that there is, unless a later theory or observation disproves it.
The basic hypothesis of GWOPM is that the banking crisis was caused by:
1) Moral hazard hypothesis: a long history of the US government bailing out financial firms. This lead to banks being more relaxed about risk taking. This made the banks sure that if their risky actions caused loss the government will bail them out.
2) Market distortions hypothesis: US government’s push to make poor people buy houses which they could not afford.
Russ Roberts provides a very journalistic formulation of the moral hazard hypothesis in the context of the banking crisis. Not a single mathematical model from information economics is mentioned. Only a superficial description of the link between government bailing out financial firms and banks becoming greater risk takers is given. This means that the moral hazard hypothesis is not testable against data because there is nothing to test. Whatever data which is then provided cannot falsify the hypothesis.
For example, Russ Roberts goes on to describe bail outs in 1980s and 1990s, again in a journalistic way. But how much influence, if at all, did this have on the behaviours of the banks? Russ Roberts thinks that it played a great influence. It may have, but it may not have. Maybe the bankers of 2000s paid no attention to what happened in 1980s because the whole of the executive branch had changed from Ronald Reagan to George W Bush and George W Bush had no incentive to follow the example of Ronald Reagan’s administration.
The market distortion hypothesis on the surface seems to be better presented. However it still suffers from the same problems as the moral hazard hypothesis. Although the links between varies financial institutions like the banks, the home buyers, mortgage companies and Freddie and Fanny are described more fully the links are still not quantified. Thus the very complicated interactions between all of these economic agents are not obvious and not testable. For example, I think that in the last 30 years there has been a cultural change from high saving and low consumption to low saving and high consumption. I think that this cultural change was not brought about by government action but rather by change in technology. Technology has created cheap consumer goods which are constantly improving in performance but are cheap . So consumers now buy a good for only a year or two before disposing of that good and getting a newer version. I think that this fundamental change in culture has played just as bigger role in the economic crisis as did government action. However without a precise formulation and defining how to test this hypothesis I face the same problem as Russ Roberts. My hypothesis may be true, after all electronics nowadays do become obsolete in 24 months, but someone could just as easy say that I am not right, and no one can determine the who is right.
Economists use methods of econometrics to test their models. Econometrics is still a very young sub-field and thus a lot of econometric findings could be argued to be false. However econometrics is the best method we have of testing our conjectures (outside of game theory which can be easily tested in a lab with real people playing games). Russ Roberts did not do any econometric testing. what this means is that GWOPM is at best only a guess.
The economy is a complex system. Most interactions are non-linear. One event causes many effects which in turn have other effects and feedback loops. This complexity is very hard to model and impossible to understand fully unless it is expressed in a very rigorous way, i.e. mathematically. Although Russ Roberts highlighted a few possible causes of the banking crisis he certainly has not proved his hypothesis. GWOPM is an interesting journalistic piece. It may serve as a starting point for a proper study of the banking crisis. However GWOPM is not a definitive account of the banking crisis. I think a much more precise and thus more compelling argument could have been made but was not.
We are still left with the question "do markets , especially capital markets, naturally spin out of control and crash, or are they naturally stable and any instability is caused by government interventions?". Russ Roberts and I think that markets work others do no, but the question is still to answered.
I am sympathetic to Russ Roberts’ motivation. Russ Roberts is a classical liberal and I share many political views with him. I think Russ and I would agree that in order to promote classical liberalism we must have good arguments for it. I think that this entails that any argument we make for classical liberalism must pass the most stringent tests available. Russ Roberts has a PhD and is a professor of economics so some may say that I am out of my league to argument with him. But this is not about a conflict between me and Russ Roberts, but rather a conflict of ideas (ideas of how to best express an understanding of the banking crisis). Because this is a battle of ideas it makes no difference who is presenting them. Ideas fight in an intellectual dimension and we are all just observers.
I think that GWOPM has many flaws. I think that overall GWOPM thus does not serve the classical liberal cause very well. Firstly I will outline on what basis I made this judgement. Then I will show some evidence to support my judgement.
Any theory is either disproved or reinforced with close scrutiny. Scrutinising of a theory is done in two ways.
Firstly the theory must be checked to be logically valid and verifiable. A theory is valid if the theory’s conclusion logically follows from the theory’s premises. A theory is verifiability only if it can be falsified. This means that the premises or the conclusion of the theory is measurable or quantifiable. This implies that theory must be formulated in a very precise way so as to be able to quantify varies effects and their causes. Otherwise it would be impossible to judge its logical validity or its verifiability.
The second way a theory can be scrutinised is to actually measure the premises and the conclusion in the real world. A theory can be logically valid and verifiable but disproved by real life data.
A theory can never be positively said to have been proven by real life observation (see David Hume’s problem of induction and Karl Popper’s Conjectures and Refutations). It may be only falsified or not falsified. However once a theory has withstood many tests against real world observations it is deemed to be the closest to truth that there is, unless a later theory or observation disproves it.
The basic hypothesis of GWOPM is that the banking crisis was caused by:
1) Moral hazard hypothesis: a long history of the US government bailing out financial firms. This lead to banks being more relaxed about risk taking. This made the banks sure that if their risky actions caused loss the government will bail them out.
2) Market distortions hypothesis: US government’s push to make poor people buy houses which they could not afford.
Russ Roberts provides a very journalistic formulation of the moral hazard hypothesis in the context of the banking crisis. Not a single mathematical model from information economics is mentioned. Only a superficial description of the link between government bailing out financial firms and banks becoming greater risk takers is given. This means that the moral hazard hypothesis is not testable against data because there is nothing to test. Whatever data which is then provided cannot falsify the hypothesis.
For example, Russ Roberts goes on to describe bail outs in 1980s and 1990s, again in a journalistic way. But how much influence, if at all, did this have on the behaviours of the banks? Russ Roberts thinks that it played a great influence. It may have, but it may not have. Maybe the bankers of 2000s paid no attention to what happened in 1980s because the whole of the executive branch had changed from Ronald Reagan to George W Bush and George W Bush had no incentive to follow the example of Ronald Reagan’s administration.
The market distortion hypothesis on the surface seems to be better presented. However it still suffers from the same problems as the moral hazard hypothesis. Although the links between varies financial institutions like the banks, the home buyers, mortgage companies and Freddie and Fanny are described more fully the links are still not quantified. Thus the very complicated interactions between all of these economic agents are not obvious and not testable. For example, I think that in the last 30 years there has been a cultural change from high saving and low consumption to low saving and high consumption. I think that this cultural change was not brought about by government action but rather by change in technology. Technology has created cheap consumer goods which are constantly improving in performance but are cheap . So consumers now buy a good for only a year or two before disposing of that good and getting a newer version. I think that this fundamental change in culture has played just as bigger role in the economic crisis as did government action. However without a precise formulation and defining how to test this hypothesis I face the same problem as Russ Roberts. My hypothesis may be true, after all electronics nowadays do become obsolete in 24 months, but someone could just as easy say that I am not right, and no one can determine the who is right.
Economists use methods of econometrics to test their models. Econometrics is still a very young sub-field and thus a lot of econometric findings could be argued to be false. However econometrics is the best method we have of testing our conjectures (outside of game theory which can be easily tested in a lab with real people playing games). Russ Roberts did not do any econometric testing. what this means is that GWOPM is at best only a guess.
The economy is a complex system. Most interactions are non-linear. One event causes many effects which in turn have other effects and feedback loops. This complexity is very hard to model and impossible to understand fully unless it is expressed in a very rigorous way, i.e. mathematically. Although Russ Roberts highlighted a few possible causes of the banking crisis he certainly has not proved his hypothesis. GWOPM is an interesting journalistic piece. It may serve as a starting point for a proper study of the banking crisis. However GWOPM is not a definitive account of the banking crisis. I think a much more precise and thus more compelling argument could have been made but was not.
We are still left with the question "do markets , especially capital markets, naturally spin out of control and crash, or are they naturally stable and any instability is caused by government interventions?". Russ Roberts and I think that markets work others do no, but the question is still to answered.
No comments:
Post a Comment