02 October 2012

Ortega y Gasset - "The Greatest Danger: the State"

from The Revolt of the Masses, chapter 13:

though it is not impossible that the prestige of violence as a cynically established rule has entered on its decline, we shall still continue under that rule, though in another form. I refer to the gravest danger now threatening European civilisation. Like all other dangers that threaten it, this one is born of civilisation itself. More than that, it constitutes one of its glories: it is the State as we know it today. ...

Early capitalism and its industrial organisations, in which the new, rationalised technique triumphs for the first time, had brought about a commencement of increase in society. A new social class appeared, greater in numbers and power than the pre-existing: the middle class. This astute middle class possessed one thing, above and before all: talent, practical talent. It knew how to organise and discipline, how to give continuity and consistency to its efforts. In the midst of it, as in an ocean, the "ship of State" sailed its hazardous course. The ship of State is a metaphor re-invented by the bourgeoisie, which felt itself oceanic, omnipotent, pregnant with storms. ...

But with the Revolution the middle class took possession of public power and applied their undeniable qualities to the State, and in little more than a generation created a powerful State, which brought revolutions to an end. Since 1848, that is to say, since the beginning of the second generation of bourgeois governments, there have been no genuine revolutions in Europe. Not assuredly because there were no motives for them, but because there were no means. Public power was brought to the level of social power. Good-bye for ever to Revolutions! The only thing now possible in Europe is their opposite: the coup d'etat. Everything which in following years tried to look like a revolution was only a coup d'etat in disguise. 

In our days the State has come to be a formidable machine which works in marvellous fashion; of wonderful efficiency by reason of the quantity and precision of its means. Once it is set up in the midst of society, it is enough to touch a button for its enormous levers to start working and exercise their overwhelming power on any portion whatever of the social framework.  

The contemporary State is the easiest seen and best-known product of civilisation. And it is an interesting revelation when one takes note of the attitude that mass-man adopts before it. He sees it, admires it, knows that there it is, safeguarding his existence; but he is not conscious of the fact that it is a human creation invented by certain men and upheld by certain virtues and fundamental qualities which the men of yesterday had and which may vanish into air tomorrow. Furthermore, the mass-man sees in the State an anonymous power, and feeling himself, like it, anonymous, he believes that the State is something of his own. Suppose that in the public life of a country some difficulty, conflict, or problem presents itself, the mass-man will tend to demand that the State intervene immediately and undertake a solution directly with its immense and unassailable resources.  

This is the gravest danger that today threatens civilisation: State intervention; the absorption of all spontaneous social effort by the State, that is to say, of spontaneous historical action, which in the long run sustains, nourishes, and impels human destinies. When the mass suffers any ill-fortune or simply feels some strong appetite, its great temptation is that permanent, sure possibility of obtaining everything- without effort, struggle, doubt, or risk- merely by touching a button and setting the mighty machine in motion. The mass says to itself, "L'Etat, c'est moi," which is a complete mistake. The State is the mass only in the sense in which it can be said of two men that they are identical because neither of them is named John. The contemporary State and the mass coincide only in being anonymous. But the mass-man does in fact believe that he is the State, and he will tend more and more to set its machinery working on whatsoever pretext, to crush beneath it any creative minority which disturbs it- disturbs it in any order of things: in politics, in ideas, in industry.  

The result of this tendency will be fatal. Spontaneous social action will be broken up over and over again by State intervention; no new seed will be able to fructify. Society will have to live for the State, man for the governmental machine. And as, after all, it is only a machine whose existence and maintenance depend on the vital supports around it, the State, after sucking out the very marrow of society, will be left bloodless, a skeleton, dead with that rusty death of machinery, more gruesome than the death of a living organism.  ...

Already in the times of the Antonines (IInd Century), the State overbears society with its anti-vital supremacy. Society begins to be enslaved, to be unable to live except in the service of the State. The whole of life is bureaucratised. What results? The bureaucratisation of life brings about its absolute decay in all orders. Wealth diminishes, births are few. Then the State, in order to attend to its own needs, forces on still more the bureaucratisation of human existence. This bureaucratisation to the second power is the militarisation of society. The State's most urgent need is its apparatus of war, its army. Before all the State is the producer of security (that security, be it remembered, of which the mass-man is born). Hence, above all, an army. ...

Is the paradoxical, tragic process of Statism now realised? Society, that it may live better, creates the State as an instrument. Then the State gets the upper hand and society has to begin to live for the State. But for all that the State is still composed of the members of that society. ... This is what State intervention leads to: the people are converted into fuel to feed the mere machine which is the State. The skeleton eats up the flesh around it. The scaffolding becomes the owner and tenant of the house.   When this is realised, it rather confounds one to hear Mussolini heralding as an astounding discovery just made in Italy, the formula: "All for the State; nothing outside the State; nothing against the State." This alone would suffice to reveal in Fascism a typical movement of mass-men. Mussolini found a State admirably built up- not by him, but precisely by the ideas and the forces he is combating: by liberal democracy. ...

Statism is the higher form taken by violence and direct action when these are set up as standards. Through and by means of the State, the anonymous machine, the masses act for themselves. The nations of Europe have before them a period of great difficulties in their internal life, supremely arduous problems of law, economics, and public order. Can we help feeling that under the rule of the masses the State will endeavour to crush the independence of the individual and the group, and thus definitely spoil the harvest of the future?  

A concrete example of this mechanism is found in one of the most alarming phenomena of the last thirty years: the enormous increase in the police force of all countries. The increase of population has inevitably rendered it necessary. However accustomed we may be to it, the terrible paradox should not escape our minds that the population of a great modern city, in order to move about peaceably and attend to its business, necessarily requires a police force to regulate the circulation. But it is foolishness for the party of "law and order" to imagine that these "forces of public authority" created to preserve order are always going to be content to preserve the order that that party desires. Inevitably they will end by themselves defining and deciding on the order they are going to impose- which, naturally, will be that which suits them best.

It might be well to take advantage of our touching on this matter to observe the different reaction to a public need manifested by different types of society. When, about 1800, the new industry began to create a type of man- the industrial worker- more criminally inclined than traditional types, France hastened to create a numerous police force. Towards 1810 there occurs in England, for the same reasons, an increase in criminality, and the English suddenly realise that they have no police. The Conservatives are in power. What will they do? Will they establish a police force? Nothing of the kind. They prefer to put up with crime, as well as they can. "People are content to let disorder alone, considering it the price they pay for liberty." "In Paris," writes John William Ward, "they have an admirable police force, but they pay dear for its advantages. I prefer to see, every three or four years, half a dozen people getting their throats cut in the Ratcliffe Road, than to have to submit to domiciliary visits, to spying, and to all the machinations of Fouche."   Here we have two opposite ideas of the State. The Englishman demands that the State should have limits set to it. 





19 September 2012

Learning Bayesian Probability Theory

Alex asked me about some introductory references on Bayesian probability theory. Here's my list (to be read in this order):

1) a historic intro:
Jaynes, 1986, "Bayesian Methods: General Background"

2) intro to Bayesian probability (and the connection to Boolean logic):
first two chapters in the book: Jaynes, 2003, Probability Theory: The Logic of Science
[the free alternative: chapters 3-6 from this Probability Theory With Applications in Science and Engineering (1973)]

3) intro on maximum entropy:
Jaynes, 1982, "On the Rationale of Maximum-Entropy Methods"
Gull, 1988, "Bayesian Inductive Inference and Maximum Entropy"

4) a general overview of the entire theory (including maximum entropy):
Jaynes, 1988, "How Does the Brain Do Plausible Reasoning?"

5) model selection and parameter estimation:
Bretthorst, 1996, "An introduction to model selection using probability theory as logic"
Bretthorst, 1989, "An Introduction to Parameter Estimation"
[the respective chapters in Jaynes' 2003 book]
[the advanced and more philosophical by the end: Jaynes, 1985, "Entropy and Search-Theory"]

6) on the connection between Bayes formula and Maximum entropy:
Caticha, 2003, "Relative Entropy and Inductive Inference"
[Caticha & Giffin, 2006, "Updating Probabilities"]

7) Bayesian theory of surprise:
Itti & Baldi, 2005-2006

8) Bayesian networks:
Jensen & Nielsen, 2007, Bayesian Networks and Decision Graphs

05 August 2012

A taxonomy of uncertainty, "WARNING: Physics Envy May Be Hazardous To Your Wealth!"

From a paper by Andrew W. Lo and Mark T. Mueller, "WARNING: Physics Envy May Be Hazardous To Your Wealth!":

Level 1: Complete Certainty

Level 2: Risk without Uncertainty

randomness governed by a known probability distribution for a completely known set of outcomes

Level 3: Fully Reducible Uncertainty

risk with a degree of uncertainty, an uncertainty due to unknown probabilities for a fully enumerated set of outcomes that we presume are still completely known

Level 4: Partially Reducible Uncertainty

situations in which there is a limit to what we can deduce about the underlying phenomena generating the data. Examples include data-generating processes that exhibit: (1) stochastic or time-varying parameters that vary too frequently to be estimated accurately; (2) nonlinearities too complex to be captured by existing models, techniques, and datasets; (3) non-stationarities and non-ergodicities that render useless the Law of Large Numbers, Central Limit Theorem, and other methods of statistical inference and approximation; and (4) the dependence on relevant but unknown and unknowable conditioning information. ...

Under partially reducible uncertainty, we are in a casino that may or may not be honest, and the rules tend to change from time to time without notice. In this situation, classical statistics may not be as useful as a Bayesian perspective, in which probabilities are no longer tied to relative frequencies of repeated trials, but now represent degrees of belief. Using Bayesian methods, we have a framework and lexicon with which partial knowledge, prior information, and learning can be represented more formally. ...

At this level of uncertainty, modeling philosophies and objectives in economics and finance begin to deviate significantly from those of the physical sciences. Physicists believe in the existence of fundamental laws, either implicitly or explicitly, and this belief is often accompanied by a reductionist philosophy that seeks the fewest and simplest building blocks from which a single theory can be built. Even in physics, this is an over-simplification, as one era’s “fundamental laws” eventually reach the boundaries of their domains of validity, only to be supplanted and encompassed by the next era’s “fundamental laws”. ...

It is difficult to argue that economists should have the same faith in a fundamental and reductionist program for a description of financial markets (although such faith does persist in some, a manifestation of physics envy). Markets are tools developed by humans for accomplishing certain tasks—not immutable laws of Nature—and are therefore subject to all the vicissitudes and frailties of human behavior. While behavioral regularities do exist, and can be captured to some degree by quantitative methods, they do not exhibit the same level of certainty and predictability as physical laws. Accordingly, model-building in the social sciences should be much less informed by mathematical aesthetics, and much more by pragmatism in the face of partially reducible uncertainty. We must resign ourselves to models with stochastic parameters or multiple regimes that may not embody universal truth, but are merely useful, i.e., they summarize some coarse-grained features of highly complex datasets.

While physicists make such compromises routinely, they rarely need to venture down to Level 4, given the predictive power of the vast majority of their models. In this respect, economics may have more in common with biology than physics. As the great mathematician and physicist John von Neumann observed, “If people do not believe that mathematics is simple, it is only because they do not realize how complicated life is”.

Level 5: Irreducible Uncertainty

Irreducible uncertainty is the polite term for a state of total ignorance; ignorance that cannot be remedied by collecting more data, using more sophisticated methods of statistical inference or more powerful computers, or thinking harder and smarter. Such uncertainty is beyond the reach of probabilistic reasoning, statistical inference, and any meaningful quantification. This type of uncertainty is the domain of philosophers and religious leaders, who focus on not only the unknown, but the unknowable.

Stated in such stark terms, irreducible uncertainty seems more likely to be the exception rather than the rule. After all, what kinds of phenomena are completely impervious to quantitative analysis, other than the deepest theological conundrums? The usefulness of this concept is precisely in its extremity. By defining a category of uncertainty that cannot be reduced to any quantifiable risk—essentially an admission of intellectual defeat—we force ourselves to stretch our imaginations to their absolute limits before relegating any phenomenon to this level.

Level ∞ : Zen Uncertainty

Attempts to understand uncertainty are mere illusions; there is only suffering.

(HT Steven Strogatz)

07 July 2012

This is an amazing book about people organizing to solve common problems


Institutions, incentives, and irrigation in Nepal by Paul Benjamin, Wai Fung Lam, Elinor Ostrom, and Ganesh Shivakoti:

There may be in excess of 100,000 distinct farmer-managed irrigation systems in Nepal, each with their own fascinating history of how farmers cooperated to develop a common resource that would benefit them all. … Farmers have identified sources of water and irrigable farmland, found the capital and companions to do the work, and constructed the systems. They and their descendants have governed and operated them. Farmers have often assumed considerable risk in irrigation investments, not just in the usual way we view risk in investment--where you could loose your shirt--but where you could also loose your life. Cutting canals along a steep slope from the source to the planned service area is not easy and lives have been lost. …

For centuries, Nepali farmers have developed their own knowledge and shaped and reshaped their mountainous and often perilous terrain. Leveling paddy fields on steep slopes, making bunds, constructing headworks, building irrigation canals and ditches, setting and adjusting field canals--all this does not just happen. Individuals conceptualize possibilities; they talk about their ideas; they decide what to do first and who should do what; they argue, have conflict, and settle disputes; they build and they rebuild; they cope with floods, landslides, and droughts; and in the process they create physical and social artifacts.

  Social artifacts are as necessary for the construction of physical artifacts (the paddy fields, the headworks, the canals, etc.) as the sticks, stones, earth, and, more recently, cement and gabion wire they use. Social artifacts are the rules and principles they use to organize their going concerns. This social capital is productive capacity that is created by the application of these rules and principles among individuals who know how to make them work. … Anyone who has bothered to take their eyes off the gleaming Himalayas will be impressed with the incredible amount of work that has gone into the construction of so much on the hillsides of Nepal. The social artifacts that stand behind that construction, however, are not so easy to see. One may see farmers at work, but not the social capital that helps them coordinate what they do. One must search more deeply to reveal the underlying social structures so essential to the accomplishment of their tasks. These are, in turn, even more deeply embedded in the accumulated knowledge organized and transmitted through language and culture giving expression to the fundamental complexities with which the farmers work.


12 April 2012

Wittgenstein on the meaning of life

We feel that even if all possible scientific questions be answered, the problems of life have still not been touched at all. Of course there is then no question left, and just this is the answer. (Tractatus)

I think this is correct, although the way he put it sounds awfully mysterious. It simply means that "the problems of life" are solved by making choices rather than discoveries. The meaning of life is about how to change things, rather than merely about figuring out how things already are. The more controversial implication of what Wittgenstein is saying is that there can be no recipe for creativity.

26 March 2012

Dan Klein, Knowledge and Coordination

Some stuff from Daniel Klein's new book Knowledge and Coordination.
p. 142-3:
"a number of things associated with entrepreneurship: 
  • Discovering (and seizing) a betterment opportunity
  • Creating new products, methods and improvements
  • Bearing uncertainty
  • Owning capital
  • Creating and/or owning a business enterprise, especially as owner-manager
  • Organizing and leading cooperation by engaging in speech, rhetoric and persuation
All these make sense and tend to travel together."
p. 141:
"Risk resides in settled interpretation, uncertainty in unsettled interpretation. Opportunities based on mere risk will be easy to communicate, to imitate - they will be obvious - and, in a free-entry environment, they will be quickly captured and closed. 'Risk' seems to imply common knowledge, like our odds with a fair coin. But 'epiphany' [entrepreneurial discovery] is typically the seeing of a new and better interpretation of things. Rather than common knowledge, it hinges on asymmetric information. A new, fruitful interpretation will be nonobvious to the extent that uncertainty, as opposed to mere risk, surrounds the circumstances and any opportunity lying therein."
Chapters 4-6
p. 37-8:
  • Beginning in the 1880s, coordination was used to describe the effective arrangement of factors and activities withing a firm. That array of factors and activities is, here, the referent concatenation.
  • In the early twentieth century, coordination was extended beyond the firm, to describe the pleasing arrangement of activities within the entire economic system, like Smith's discussion of all that goes in the making of the woolen coat. Here, the referent concatenation would be a much wider skein of factors and activities, even the global economy.
  • Beginning around 1960s, with Thomas Schelling and game models, coordination took on another meaning: mutual meshing of actions, one to another. In the ensuring decades, the mutual idea of coordination caught on and became focal.
  • Nowadays, the idea of mutual coordination overshadows concatenate coordination. Also, the two are sometimes conflated.
p. 60-2: the explanation of the connection between the two meanings of coordination and the distinction between spontaneous order and emergent conventions. Chapter 6: confusions related to mixing up the two meanings of coordination.
p. 62:
Concatenate coordination (Smith, Hayek, Coase) -> spontaneous order
Mutual coordination (Schelling) -> emergent rules
An arrangement or concatenation pleasing to beholder or spectator"Meshing" interaction within a situation of coincidence of interest
In some cases, the concatenation and its coordination are only abstract or notional from the interactor's point of view ("invisible hand")Manifest from the interactor's point of view
A broad topic that includes features often out of step with with the rubric of mutual coordination, features such as disjointed knowledge, asymmetric interpretation, discovery, innovation, entrepreneurship, competition, abstention, exit, shunning, and bargaining.Of a rubric that includes: focal points, coordination equilibrium, coordination games, convention.
p. 12: "I call respondence our rather automatic responding to new bits of information that simply rain down on us. Respondence is like search, except that you weren't actually searching for the information acquired. Search is active and costly, while respondence is passive."
p. 13: "It is one thing for the entrepreneur to greet fortune when it comes knocking. It is something else to apprehend fortune in its hidden forms and seize it. Here, we have the distinction between responding to the realization of events within a framework of recognized variables and relationships and the discovery of fresh opportunity to embrace a new and better framework or interpretation. This element of epiphany, of finding fortune by interpreting the world differently, is the subtle and vital element in human decision making. Yet it is absent from equilibrium model building. In equilibrium stories, agents never have a 'light bulb' moment."
p. 15: quotes Kirzner (1985): "To Kirzner, the most impressive aspect of the free enterprise system is not its ability to generate efficient allocations within a framework of fully recognized ends and means. Rather, 'the most impressive aspect of the free market system is the tendency for [previously unrecognized ends and means] to be discovered' (30)."
Serendipity: p. 17: quotes Alchian: "The survivors in the market, he explains, 'may appear to be those having adapted to the environment, whereas the truth may well be that the environment has adopted them' (1950: 22)." 
p. 19:
Static choiceRespondenceEpiphanySerendipity
Max U.switchboard operatorthe entrepreneur (the verger)
no discoverysmall discovery within one's interpretative framework discovery that alters one's interpretative framework<- idem
more deliberate

less deliberate

27 January 2012

Filip Palda's "Pareto's Republic and the New Science of Peace"

The PPE workshop today had Filip Palada presenting his book Pareto's Republic and the New Science of Peace. Palada's main concern is with the process that creates more efficient institutions and he notes that it's not an accident that better institutions seem to be associated with more peaceful states of affairs (not just less war but also less private violence). He noted that secure property rights in an evironment of rule of law are key, but he didn't give an answer as to what factors lead to such a system. Here are some of my thoughts on it.

First of all, Peter Leesom is right that the book should be called The Kaldor-Hicks Republic, as that is the concept of efficiency he seems to be using when he's assessing whether a set of institutions is better than another. This detail aside...

There were lots of discussions about how dictators and their cronies could be compensated such that they would agree to leave (compensation + a changed regime being better than no compensation + the regime). Peter Boettke mentioned that this is virtually impossible to do due to the problem of credible commitment: How would you assure the dictators and their cronies that they would really get to keep the promised goodies after they agree to leave? The amount of money necessary to compensate them for this risk is simply too much. So, the alternative is similar to what has happened in post-communist transitions: the former nomenklatura still maintains some power and the wealth that comes with it.

Of course there's also a different kind of compensation possible: You can tell them, "if you agree to leave, we agree not to kill you". I assume Gadhafi would have been wise to take such a bargain while it was still possible. This hints to what I think is missing in Palada's approach: what ultimately determines the changes in institutions is the power struggle between different actors. Formal institutions can be seen as tools used by those with a position of power to try to secure their position and the associated rents. Culture, while posing some constraints (how much bullying are people willing to take?), is not that important because people adapt to whatever the institutional environment happens to be, i.e. culture changes to reflect what's prudent to do at the individual level within the given institutional setting (more or less e.g. why don't North Koreans have more of a Guns of Brixton mentality, while the Tibetans seem to do). So, if you want to understand the way institutions change you have to track the power struggles and what factors (such as technology or international forces) change the balances of power.

It is often said that violence should be the means of last resort. For example, in the Revolt of the Masses Ortega y Gasset defined the degree of civilization of a society as the number of options or methods available in that society for solving conflicts before resorting to violence. In practice however, violence is usually the means of first resort. The microeconomics of this is something along those lines: if you are much more powerful than me, you face a very strong temptation to just take my stuff or bully me into some sort of exchange that is disadvantageous for me (and with which I wouldn't have agreed under non-threatening conditions). Morality may offer some restrained to the possible bully, but this is not a reliable long-term solution because the immoral have larger gains compared to the moral and thus gradually take over.

Consequently, an environment of (quasi-)voluntary exchanges is the result of a balance of power. Olson's story about the fact that stationary bandits (aka governments) win over roving bandits is along these lines, and this story seems to have a pretty good historical basis (see the 'From Egalitarism to Kleptocracy' chapter in Guns, Germs, and Steel).

I would say that it is even impossible to define what a voluntary exchange is, if the two agents are not of similar power. This is because violence has two uses: (1) one is predation of the powerful over the weak, (2) the other is enforcement of contracts and agreements. If you do me a service based on my promise that I would give you some money (or service) at a later date, what prevents me from abandoning my promise is the credible threat of what would happen to me if I don't deliver my part of the deal. But that threat is credible only if it is made by someone sufficiently more powerful than me (or by the possibility of lost future revenues via some reputation mechanism - like the eBay or Amazon ratings systems). The problem is that, if the deal was made with such a much more powerful agent, what assures us that the deal really was voluntary? After all, it would have been awfully generous of such an agent to restrain himself and not bully me into “agreeing” to a disadvantageous exchange. In other words, if the agreement is between two agents of very unequal power, there is no way to tell whether the powerful agent uses violence for predation or for enforcement. Thus, under such conditions, the very distinction between voluntary and involuntary agreements disappears (at least if you agree that a distinction exists only if it can be measured).

This means that we can talk of voluntary exchanges only when the two parties are of similar power, and there is a third party enforcer (either an agent such as a government or a network such as the reputation networks on eBay or Amazon) who is much stronger than the two. This also means that if the third party enforcer is a specified agent, there can be no real voluntary agreements between a normal agent and this much stronger enforcer. I.e. there is no such thing as a voluntary contract with the government (because there's nothing you can do if the government doesn't comply with its contractual obligations - see the current "debts restructuring" discussions in which private banks lose tons of money as states no longer pay them back - and nothing you can do if the government decides to attach some obligations to you that you don't really like and for which you are not compensated - e.g. mandated health insurance). It also means that there can be no such thing as anarcho-capitalism as it is normally construed – i.e. assuming contracts between normal people and protection agencies much stronger than them (because he who has the power to protect you, also has the power to abuse you, regardless of whether that “protector” is state or private).

So, to answer Palada's dilemma about what can be done to improve the institutional settings and increase peace, we need to focus on the way various power centers constrain each other. This is of course a very old and good idea: the concept of checks-and-balances and of federalism. But what factors lead to better checks and balances and create a larger safe space in which voluntary agreements can happen?

One way to think about it, not necessarily the best, is the Rothbardian theory of history, which assumes there are cycles of power losses due to authorities being taken by surprise by certain technological inventions that (temporarily) empower the weak. E.g. the invention of the printing press undermined the power structure, but eventually the invention was captured by government – the modern bureaucratic state would not have been possible without printing. The industrial revolution and the emergence of cities undermined the power of the land-owning aristocracy, but eventually capitalism morphed into the modern neomercantilist system as the new capitalist elites found the state useful as well. Radio was quickly captured by the state as the EM spectrum was nationalized, and it was effectively used for propaganda by the Nazi and fascist regimes. Now the internet has taken governments by surprise, and they are currently working hard to get back some sort of control (e.g. think SOPA/PIPA and ACTA).

So, from this perspective, the only truly exogenous factor is technology (in the sense that it generates surprises with wide-spread and unpredictable consequences for the existing power structure). Institutions and culture are endogenous: institutions are tools used by the current elites for (attempting) to maintain control and preserve their rents, and culture is more of an effect than a cause (i.e. culture understood from an act utilitarian point of view, and being driven by whatever is prudent for individuals to do in a given context, while somewhat constrained by our biological moral sentiments, such altruistic punishment etc.).