To better understand the failure of Western democracies on the issue of economic advice, one can best start by reading Paul Krugman’s books “The Age of Diminished Expectations” (1990), “Peddling prosperity” (1994), “Pop Internationalism” (1996), and “The accidental theorist” (1997). For example, when Krugman discusses US majority leader Armey’s book “The Freedom Revolution”, he states: “Armey is no fool. He cannot be unaware that he is fudging his numbers. Possibly he regards a small fib as justifiable in the service of a higher truth. Or possibly he has managed to achieve a state of doublethink, in which the distinction between what is politically convenient to believe and the objective facts no longer exists [sic]. The end result is the same: His book is an effort to obscure the stark realities (…)” (1997:60). Similarly, one can read in the American Economic Review that the US Council of Economic Advisers is rather proud of its achievements in the last decades, but we should be aware that this council is a bureaucratic body, and it hasn’t the independent position that could have protected the US economy from the events and errors as are related by Krugman in his “Peddling prosperty” saga or shown by the record of mass unemployment.
Let us now regard what the West could have done with regards to Russia after the fall of the Berlin Wall and the first free elections there - and what could be done now also with respect to Asia. I take my own 1996 paper “Enable Russia to help itself”, and quote from its summary: “Western nations in the 1990s hinder trade with Russia and the Eastern nations for fear of unemployment at home, as they did in the 1930s with Germany. If trade were stimulated instead of hindered, Russia could regain economic and political stability by itself. The moral problem is not external and does not concern whether Russia would need financial aid. The moral problem is internal, and concerns whether Western political leaders are willing to face their own errors that cause the present mass unemployment at home.”
Clearly, with this being the state of affairs, one can imagine the strength of the forces that prevent a proper discussion of these issues. Western companies embrace tariff barriers to cheap imports - and raise their own prices. Bureaucrats embrace barriers since these give a sense of control, and these also justify the very existence of this bureaucracy. Labour unions will fight unemployment at home with whatever misguided argument it takes. Governments embrace economic tales about ‘globalisation’ and ‘competition from cheap labour countries’ since these distract attention from home grown errors, and these goverments neglect economists who tell them that ‘globalisation’ and ‘competition from cheap labour countries’ are rather like fairy tales indeed. Krugman again uses the term ‘globaloney’ - and have you heard your President or Prime Minister adopting that critical attitude too ?
The best economic advice for the current situation is as follows - and I urge upon my fellow economists to adopt and spread that advice too: Every parliament could install a committee that will enquire into the process of economic advice. This committee could study Krugman’s books and my suggestions for a solution of mass unemployment and for an Economic Surpreme Court amendment to the national constitution(s). Nothing less will do. Note, by the way, that when countries start installing these committees, the markets will be quick to anticipate the directions of their conclusions, and economic recovery would already set in.
We all know Lincoln’s words: “You can fool all of the people some of the time, and you can fool some of the people all of the time, but you cannot fool all of the people all of the time.” Let us act upon it, or show Lincoln wrong. (August 1998)
Notes in 1999: (1) A 1999 UNDP report describes the Eastern European situation as disastrous, and calls for a quick joining up to the EU (De Volkskrant October 16 1999). It is courageous that an international body speaks up like this - and it indicates the seriousness of the situation. (2) The journalist Peter Michielsen in NRC-Handelsblad October 30 1999 rightly calls attention to the original borders between the empires of Rome and Byzantium. The Eastern European countries that are doing relatively well belong to the Roman area, the others to Byzantium. He mentions that this cultural distinction has also been noted by Andreas Oplatka of the Neue Zürcher Zeitung 1994, who again refers to George Kennan in 1945. I was a bit surprised by this, hadn’t thought about it in this way. (3) These points however nicely fit what I have been argueing for ten years now. Enabling people to help themselves starts with taking account of the local conditions; and overall the barriers to trade should go.
At the Dutch Central Planning Bureau, I helped making the Athena model (CPB (1990)) with its 7000 variables. I had this model at my computer and could let it do tricks like an obedient dog. But a proposal to an exercise effectively like the above was rejected by the directorate, and nowadays I am no longer in the position to make such proposals. The desktop computer that I have now, in 2004, might have more power than the 1990 mainframe, but I don’t have the data, the programs, and the possibility of discussion with colleagues. I have Word for Windows, Mathematica, some crucial books, an occasional visit to the Dutch Royal Library, and the internet (at low speed). Moreover, I have to make a living, in a different kind of job, and my time constraints thus are severe. This explains why I am forced to a logical argument - and this explains again why I emphasise logic anyhow.
Thus, crucially: it is up to the fellow economists to check my findings. They / you should actually do this anyhow, since a critical perspective always is best. For example: What are the data on the minimum wages in the other OECD countries ? OK, the OECD internet site shows that 1997 statutory minimum wage is 39% of median wages incl. overtime in the USA, 60% in France, 30% in Japan, etcetera, quite sizable [53] - but what about the tax void, the development, the indexation, the discouraged workers below the minimum, etcetera ? [54] What about the shifts of the Phillipscurves in this light ? What about the effects of the dynamic marginal rate ? How are these topics in all nations ? And what would happen, if all nations gain confidence about growth policies again, and they fire up each other and move all to a new higher growth path ? Clearly, the research agenda is huge.
The situation since 1989-1991 has been a bit like this: Me stating that unemployment has been solved (analytically) and inviting the fellow colleagues to check it - and nothing further happening. This book should make a difference in that I collect the various articles that I have been able to write since then. When others see the whole route then they will also better see the crucial junction where to take the other turn.
This may also concern the novel contribution to methodology below. [55]
Methodology may be seen as ‘economics applied to science’. The methodology of economics is the fixed point in that construct - even economic methodology in the traditional form as presented by Tintner (1968).
The ‘basic economic problem in science’ is - in my perception or definition - that some set of concepts can better deal with the data than another set. New ideas are like manna from the sky, but the manna must be collected, stored, compared to the older findings, etcetera, and an optimum must be found, using scarce resources over alternative ends. This ‘basic economic problem in science’ thus is quite different from the ‘mundane (non-basic) economics’ that, say, 5% more truth can be traded against 10% more effort and cost.
The mind has the economic problem of dealing effectively and efficiently with (i) old concepts, (ii) new information and (iii) the construction of new concepts. The name of the game is to have concepts or definitions fit reality as usefully as possible. The definitions must be chosen as strong as possible, so that uncertainty can be shifted to observation (and the problems with observation).
The human mind seems to be occupied with reduction of cognitive dissonance - or, at least, that is a fruitful way to look at that mind. Here I follow Aronson (1992a&b), who provides a definition of cognitive dissonance, and data and tests that lend empirical support for it. It appears that a commonly used method of reduction of cognitive dissonance consists of the rejection of new information to the advantage of older views. Frequently the messenger is blamed for the bad message, and even, after the messenger has been punished, the bad news is neglected since it came from an unreliable source - namely a person who had to be punished (while it is forgotten that, if the news is considered irrelevant, then there was no base for punishment). Man is a rather prejudiced creature, and thus not so effective and efficient at information handling - but man has to handle new information.
Barrow (1998:4) [56] provides us with a useful quote:
“This unifying inclination of ours is a by-product of an important aspect of our intelligence. Indeed, it is one of the defining characteristics of our level of self-reflective intelligence. It allows us to organize knowledge into categories: to know vast numbers of thing by knowing rules and laws which apply in an infinite number of circumstances. We do not need to remember what the sum of every possible pair of numbers is: we need know only the principle of addition. The ability to seek and find common factors behind superficially dissimilar things is a prerequisite for memory and for learning from experience (rather than merely by experience). (…)
All human experience is associated with some form of editing of the full account of reality (‘we cannot bear too much reality’). Our senses prune the amount of information on offer. Our eyes are sensitive to a very narrow range of frequencies of light, our ears to a particular domain of sound levels and frequencies. If we gathered every last quantum of information about the world that impinged upon our senses they would be overwhelmed. Scarce genetic resources would be lopsidedly concentrated in information-gatherers at the expense of organs which could exploit a smaller quantity of information in order to escape from predators or to prey on sources of food. Complete environmental information would be like having a one-to-one scale map. For a map to be useful it must encapsulate and summarize the most important aspects of the terrain: it must compress information into abbreviated forms. Brains must be able to perform these abbreviations. This also requires an environment that is simple enough and displays enough order, to make this encapsulation possible over some dimensions of time and space.
Our minds do not merely gather information; they edit it and seek particular types of correlation. They have become efficient at extracting patterns in collections of information. When a pattern is recognized it enables the whole picture to be replaced by a briefer summary form which can be retrieved when required. These inclinations are helpful to us and expand our mental powers. We can retrieve the partial picture at other times and in different circumstances, imagine variations to it, extrapolate it, or just forget it. Often, great scientific achievements will be examples of one extraordinary individual’s ability to reduce a complex mass of information to a single pattern. Nor does this inclination to abbreviate stop at the door of the laboratory. Beyond the scientific realm we might understand our penchant for religious and mystical explanations of experience as another application of this faculty for editing reality down fo a few single principles which make it seem under control. All this gives rise to dichotomies. Our greatest scientific achievements spring from the most insightful and elegant reductions of the superficial complexities of Nature to reveal their underlying simplicities, while our greatest blunders often arise from the oversimplification of aspects of reality that subsequently prove to be far more complex than we realized.”
This human property should be used in economics to explain actual events. Colignatus (1996d) for example applies Aronson’s findings in social psychology to economics, trying to indicate the actual ‘forces’. Another application is the very analysis in this book, for example where we stated earlier:
“If the government on the one hand would desire to use the results of scientific advice for its budget process, and on the other hand would not opt for an Economic Supreme Court, then its definitions would be logically inconsistent, and it would thereby tend to create a cause for dishonesty and improper manoeuvreing and thereby corrupt its processes.” (above)
While the above relies on structural models, the property can also be modeled in the reduced form. Chapter 40 uses information indicator I {0, 1}.
Another application is to the methodology of science. Methodology should harness this human property, and clarify when it is useful and when it is misleading.
Science aspires at a more unbiased approach. This unbiased approach also means the deliberate creation of cognitive dissonance, by creating new concepts and by looking hard at the evidence till it doesn’t go away anymore.
The evolution of knowledge can be described in terms of an ever increasing power in the concepts used.
The introduction of a new definition is not simple. The questions always are: does the definition cover the facts as we know them, does the definition not introduce hidden aspects that cause confusion and prevent advancement ? If a new definition wins out, it is, apparently, only so because it is believed to have passed the test. Though, we should be critical of this assumption. Only if the environment is ‘critical’, then we might presume a ‘survival of the fittest’ for concepts. (And all this is reminiscent of Dawkins’s ‘memes’.)
Definitions can be devious in quite vulgar ways. In the English economics literature, ‘perfect competition’ is defined as the situation when no agent can affect the price, i.e. all agents are price takers. The Dutch word for this case is ‘full competition’. The English definition forces English economists to use the word ‘imperfection’ for all other cases. Even quite reasonable cases, in the normal state of human life, when agents have market power but balance at some social optimum, would be ‘imperfect’. Also a natural monopoly would be an imperfection - even if one could not conceive the situation differently since the monopoly is a natural one. It would be better if the English economists would adopt the Dutch definition, so that the words ‘perfect’ and ‘imperfect’ could be used in their proper sense depending upon circumstance. This is just a vulgar example of how definitions can lead one astray.
The competition of alternative concepts can be quite sophisticated however. Let us illustrate this with three examples. The most illuminating example may well be Pythagoras’s theorem and its relation to the circle. This problem concerns mathematics, so that the discussion is less taxed by semantics and empirical matters - though there is of course the theory about empirical space. The second example of ‘falsification’ is surely in the realm of empirics. The third example concerns the distinction between determinism and volition.
Regard a triangle with perpendicular sides a and b and hypotenuse c. There are two points of view:
1. Pythagoras proved [57] that the square of the hypotenuse equals the sum of squares of the perpendicular sides, i.e. that a2 + b2 = c2
2. For the circle, it is taken as the defining quality of the circle, and thus accepted without proof, that the points are at equal distances from the origin. In other words, a circle with radius c is defined as the collection of points (a, b) at a distance of c from the center. Thus a2 + b2 = c2 by definition.
The two points of view are presented in Figure 16. The definition of the circle can be taken for granted, since it is just a definition. On the other hand, it will be very useful to discuss the proof of the Pythagoras theorem, since then we see the need for a proof.
Let us take the square with sides z = a + b and surface z * z = z2 = (a + b)2. Within this square we can see four triangles with straight sides a and b and hypotenuse c, as has been done in Figure 16 in the square on the left.
In the square, another tilted square has been drawn, with sides c and thus a surface of c2. There are four surrounding triangles, each triangle has a surface of ½ a*b. The surface of the large square is equal to the surface of the tilted square and the four triangles.
Figure 16: Pythagoras and the circle
Thus:
· From the big square itself: z2 = (a + b)2
· From the tilted square and the triangles: z2 = c2 + 4 ab/2.
Elimination of z then gives a2 + b2 = c2.
This proof has been taken from DeLong (1971), and he remarks that Pythagoras proved it differently.
How do we explain that one and the same equation can have two interpretations that are so widely different, one with the need for complicated proof and the other with direct acceptance by definition ?
There may be other explanations, but I think the following will do fine. Note that the definition of the circle relies on the notion of ‘distance’. There are two points of view again, so that point 2 above actually splits in two parts:
2A) Basically the (Euclidian) distance between two points can be measured by a straight line section. That is rather simple, and makes for a readily acceptable definition of a circle.
2B) However, in a system of co-ordinates, that distance can be reinterpreted in a representation in terms of the co-ordinates. There are two possibilities again. Either the distance can be defined as simply the formula dist[{x, y}, {a, b}] ((x - a)2 + (y - b)2 ) with {x, y} the origin - above {x, y} = {0, 0} - or it can be defined geometrically as the hypotenuse of the differences of the co-ordinates. If either definition is accepted, then one can use Pythagoras’s theorem to derive the other.
The essential difference between (2A) and (2B) is that (2A) is elementary and poor in concepts and results, while (2B) is complexer and rich in concepts and results. Viewpoint (2A) only allows us to use measuring rods between arbitrary points and little else. We are allowed to sweep the rod around the center, and thereby draw the circle, but then it somehow stops. Viewpoint (2B) allows us to do much more. A line between two points is interpreted in terms of a system of co-ordinates, and that opens the scope for new results.
We find that the opposition of (1) against (2) is rather messy, and (2) actually hides two suppositions. The ease of (2) depends directly upon the ease of (2A), while (1) actually compares with (2B) that is complexer. The phrase “In other words” in (2) above thus was misleading, and actually represents the introduction of another assumption.
With this clarified, we also note that (2) is stronger than (1), and that it was possible to seduce the human mind to accept (2) rather easily. There has been a progression in concepts, resulting in stronger definitions.
Note that behind all this there is a notion of empirical space. In (1) there is a hidden assumption of a flat space. In (2B) the assumption is made explicit, and then open to amendments (curved surfaces, or abstract spaces). The movement of (1) to (2) thus is, partly, (a) the advancement in concepts by means of the definition of distance (and the circle as a collection of equal distance points), (b) the introduction of the separate step of observation - with the difficulties: when does the definition apply to reality, or if there is some reality, how do I select the proper definition ?
The point that is relevant for this book then is: that the definition is so good, that it in practice substitutes for many everyday empirical problems. A criterion for a good definition is: that it can be such a substitute.
When a definition is a close substitute for reality, then it may percolate into common culture with more authority. For example: every citizen can establish the existence of a tax void and Pareto suboptiomal unemployment purely from the logic of the level of gross minimum wages and the official tax statutes - and we don’t need big computers or official bureaus to do some econometrics and then tell us.
Admittedly, there is danger in seductive and seemingly right but wrong definitions. If ‘child’ is defined as ‘irresponsible young human’, then we may be tempted to treat children as such and forget to expect the responsibility that they can handle. But the existence of this danger should not make us close our eyes to the advantages of good definitions.
A side issue concerns our concept of ‘space’. Let us first consider an example of cultural relativism. It appears that different human cultures can have different approaches to one’s orientation in space, and that these approaches are wired into the languages used. [58] Taking a point of reference can be done in three ways: (1) Relative: taking one-self (“the tree is to the left of the house” - seen by me); (2) Absolute: taking the sun (“the tree is to the west of the house”); (3) Intrinsic: taking one of the objects (“the tree is to the back of the house”). If someone is asked to copy a situation in front of him towards a place in the back of him, then there will be a different ‘copy’ depending upon one’s language/culture. If you have a cup of coffee and a pencil in front of you, pick them up, turn yourself around, and recreate the scene, then a Westerner will use relative positions, while an Australian Aboriginal will use absolute positions (and turn the relative positions around). The question now is: while this only concerns the point of reference, can we imagine something similar that affects our concept of space itself ?
I take the position that the human mind apparently is able to conceptualise Euclidean space - and that this actually defines our concept of space. If we take a non-Euclidean geometry - such as a globe - then this still can be imagined to exist within Euclidean space. Pythagoras’s theorem is invalid for triangles drawn on a globe, but to hold that space is a globe would be erroneous - since our definition of space would be Euclidean.
One of the questions often posed is whether the universe - interstellar space - is Euclidean or not. This is a badly posed question. If we define space as Euclidean, then it is another question whether a ray of light follows a straight line or is deflected by gravity.
Barrow (1998:p42-44) provides a troubling quote: [59]
“The most important consequence of the success of Euclidean geometry was that it was believed to describe how the world was. It was neither an approximation nor a human construct. It was part of the absolute truth about things. (…) This confidence was suddenly undermined. Mathematicians discovered that Euclid’s geometry of flat surfaces was not the one and only logically consistent geometry. (…) None had the status of absolute truth. Each was appropriate for describing measurements on a different type of surface, which may or may not exist in reality. With this, the philosophical status of Euclidean geometry was undermined. It could no longer be exhibited as an example of our grasp of absolute truth. (…) These discoveries revealed the difference between mathematics and science.”
This quote is troubling for the following reasons:
1. If we define ‘space’ as Euclidean, then it is an absolute truth. This definition seems to maximise our information power. Other surfaces can be imagined within that space.
2. One might think of ‘empirical space’ as something that must be measured. The idea is: ‘If it cannot be measured, then it is not relevant.’ OK, this seems fine in principle. But if a physicist would use ‘light’ as a measuring rod, then this is asking for problems. Namely, Euclidean geometry already provides us with our system of measurement. Defining ‘empirical space’ differently would conflict with our original definitional grasp of space. Better is: to stick to the definition, and regard measurements that deviate - e.g. from gravitational deflection - as the physical properties of the objects and measurement tools involved.
3. That there is a difference between mathematics and science does not disqualify the notion of absolute truth. A true deductive sequence ‘Assumption Conclusion’ has absolute truth. And it should be realised that scientific theories are mathematical (with the scientist working on an assumption).
4. It is possible to translate the Dutch ‘lijn’ as ‘point’, and ‘punt’ as ‘line’ (thus conversely) and still find a consistent model for Euclid’s axioms. But this is a mathematical exercise, and it does not necessarily have to do with ‘space’.
So it seems that Barrow and I agree for 99%, but still, the 1% difference features big in some dimension. Note that the discussion here concerns more a side issue, but it remains useful to indicate the deeper aspects of Pythagoras’s theorem.
The ‘principle of falsification’ is that hypotheses are only scientific if they are formulated such that they are vulnerable to empirical testing, and might be falsified. It has been formulated by Popper, see Keuzenkamp (1994).
The principle has two disadvantages: (1) purely logical, (2) stochastically.
(ad 1) Take logic first.
Counterargument 1. Regard the statement All ravens are black. This statement will be false when one finds a non-black, say white, raven. So the statement would be an acceptable scientific hypothesis, since falsification is possible in principle. But, as the falsificationist would hold, it would remain a hypothesis, and we should be aware of the fact that is only a hypothesis, until it had been checked for all ravens (Tintner (1968:12)). This falsificationist view however is problematic, since most of us will sense that there is truth in All ravens are black, for example by our definition of a raven.
Counterargument 2. In the extreme, all scientific knowledge would consist of instances of falsification. It has been falsified that the Earth is flat, that atoms cannot be broken, that ... But the principle itself, i.e. that ‘all scientific knowledge would consist of instances of falsification’, is a definition and is not open to falsification.
While falsification may be a successful research strategy in many cases, it does not seem to be a fully satisfactory way of organising science, at least from these two points of logic.
(ad 2) Take stochastics next. Let us regard the typical modelling situation:
|
The model: Estimation: Observation X[+1] forecasts: Final observation: |
y = X ß + y = X b + e yest[+1] = X[+1] b + Exp[e[+1]] y[+1] |
The question now is whether this new observation can falsify the hypothesis of the empirical estimate. This question is not as simple as the naive falsificationist first had in mind. The principle of falsification is formulated as for deterministic reality, while many empirical models are stochastic. In stochastics, there may be deviations, and sometimes large ones. There are problems of measurement in y and X, the choice of the functional relationship, missing variables, and the choice of the stochastic specification itself.
One useful empirical answer is optimal control, with the example of a rocket launched to the moon, where there is continuous adjustment to observed error (‘falsification‘). This control only works well when there is a proper definition of the loss function. The issue of the loss function is a crucial one, but this is not falsificationism.
Logic and stochastics cause me to take the following position.
There is a difference between all1 (universal) and all2 (generally, usually, normally). The statement All ravens are black can be seen as:
1. a definition. It then holds universally. Empirical truth then is conditioned to the logical tautology of the definition that we have chosen. If we find a white bird that looks like a raven, it cannot be a raven. (But we think that this definition covers reality, for example since we have some ideas about genetics and evolution.)
2. an empirical statement - grounded in a stochastic model. It is shorthand for All ravenlike birds tend to be rather black or whatever the professional might deem correct. The meaning of such statements is more subject to context than in the case of well-groomed definitions.
The human mind thus faces the choice: To adopt a definition and run the risk that this does not fit reality so well, or to adopt a statement on averages and work out more details of the empirical loss function. Decisions on such statements thus are sensitive to the loss function, but the second category requires more detail.
This of course does not solve everything. The distinction of these two dimensions or perspectives is not like solving all problems in their domains. Also a definition like All ravens are black by definition does not answer the question whether a particular object is a raven or is black. Is a size of 10 kilometers acceptable ? Did we look in daytime or at night ? Must it be alive, and then, what is life ? So the distinction between definitions and empirical statements is useful, but it does not solve all problems. The point is not quite that one can always adjust definitions, but rather that a definition is not reality by itself. (Though it can get close.)
At one point in history, scientists were willing to accept the periodic system of elements to catalogue the wide variety of materials around us. There was apparently little loss involved in accepting these definitions, or Lavoisier’s periodic table was more gainful than other catalogs. The definitions did not change the materials, but facilitated more efficient research. At one point in history, see Mirowski (1989), economists were willing to analyse human behaviour in terms of utility maximisation. The approach is an empty box, since any behaviour can be described as such. For example satisficing behaviour can be represented as minimising the distance from satisfaction. Also in ‘evolutionary economics’ the utility maximisation model can be applied though these researchers are critical of this approach. (While, curiously, Charles Darwin was inspired, amongst others, by Adam Smith.) The new approach for laboratory experiments makes us even more critical about the rationality hypothesis. Utility maximisation however helps organising one’s thoughts, helps professional discussion, facilitates modelling and empirical estimation, and is generally considered an advance above less explicit approaches.
As with the Pythagoras example, but now empirically, there is a switch from just empirical knowledge to a set of definitions, when the loss function allows it.
Kuhn (1962) describes major changes as ‘paradigm switches’ (though someone noted that he used that word in perhaps 40 ways). I rather draw attention to the change from empirical knowledge to definition. This change need not be a paradigm switch. Paradigm switches may be the most intriguing or flashy examples of the introduction of new definitions, but the change from empirical knowledge to definition does also occur in ‘normal science’.
Holland around 1600 had the theological argument between Gomarus who defended predestination and Arminius who defended a measure of volition. This discussion had started before them, didn’t end with them, and continues till this day, also in these pages.
The 20th century gave a novel twist to the argument, namely quantum mechanics. Instead of the folly of the gods, there now is a randomizer with a scientific garb. If objects, and the molecules in our brains, have random aspects, then this would be neither determinism nor volition. Quantum mechanics normally is applied at the micro level of particles, and there is the suggestion that larger aggregations of masses still would behave in the Newton-Einstein fashion. Schrödinger however gave an example - his cat - how quantum mechanics could also extend into this macro world. So the challenge to the debate on predestination is real. [60]
The quantum model is stochastic of itself. This differs from the randomness caused by simple measurement errors - the randomness commonly used in economics. However, economics has some purely stochastic models of itself too. There is for example the Erlang queueing model. Consider a postoffice with clients arriving and being served. Interarrival and service times can be modeled with exponential distributions, and this allows us to determine the average length of the queue, the average waiting time, the average utilisation rate of the service window, and such. If the situation gets more complicated, then research economists use computer simulation models to find the best way of operation. This example shows that economics already is familiar with a model that is stochastic in itself. Note that there are some ways to re-introduce a degree of determinism - as your barbershop may require you to make an appointment. The basic observation that we make here is that the stochastic approach is basically a modeling method, and there is no implication that arrival and service are intrinsically random.
The discussion above introduces the various components, and the question now becomes what to make of it all. The following gives my solution.
First of all, science by definition avoids the ‘deus ex machina’ assumption. An understanding of reality is looked for without reference to a god. So our discussion is not burdened with the associations of eternal damnation (and predestination to this).
Secondly, science by definition aspires at a deterministic understanding. Scientists may adopt a stochastic approach with only a limited degree of accuracy, but the target remains a 100% accuracy - which is determinism. Hence, by definition, scientists have a deterministic predisposition. [61] [62]
Thirdly, the idea of a ‘free will’ is a moral category, differing from physics. Admittedly, the scientific approach would presuppose that our moral considerations depend on our brain, and the movements of electrons and molecules that could be caught in a determistic model - but the proper conclusion is that we don’t have that model yet. The existence of time, and in particular the uncertain future, is a precondition for morality. An ‘existence proof for God’ would be that in the limit of time, prediction accuracy rises to 100% and all moral beings are going to make the proper moral choices. [63] But we don’t know for sure that those choices will be really moral - and anyway it is hard to see how this could affect us. For example, we may predict, as social scientists, that when economic conditions worsen, that politicians then may be more inclined to morally dubious choices. But we need the passing of time to determine whether this prediction materialises - and, as human beings, we would still want to form a moral opinion and discuss the moral aspects. The conceptual gap between ‘ought’ and ‘is’ remains. Eventually there might be a practical (non-conceptual) bridge, but for those same practical reasons it isn’t there yet.
Though science does not refer to gods, we can use a god anyway for clarification. Janus, the Roman god and name-giver to the month of January, had two faces, one to the past and one to the future. Figure 17 uses the Janus head as an analogy to locate the various concepts.
Note: This only displays the three opposing concepts
in one picture,
without implying that all concepts to the left are equal
or that all concepts to the right are equal.
The Janus head analogy works only up to some degree. We don’t know all that happened in the past, we can use probability statements for the past too, and thus we cannot replace ‘past’ with ‘certainty’. Similarly, as said, science has a deterministic predisposition, so the future basically is predetermined from a scientific point of view. Yet the head analogy is useful, since it focusses our attention to these various subtleties.
Thus, clearly, the Arminius and Gomarus debate can be seen as non-sensical if they got the two categories of science and morality confused. Even though we can have a deterministic predisposition, we still can have moral volition (and be judged by jurors on making wrong choices). Their debate would be proper in so far as Gomarus would take predestination in a moral sense - but then the debate is not relevant for us.
Thus, clearly, quantum mechanics drops out as a fundamental category. It only remains as a research strategy in the face of apparent difficulties, but it still is on the road to 100% accuracy.
Admittedly, quantum mechanics itself seems to pose that nature would have random properties at the micro particle level. Some even argue that this would be the basic example of true probability - while all other ‘examples of probability’ (like throwing dice) are basically deterministic (and we only use probability techniques to make up for our lack of knowledge or laziness in measurement). In particular, Richard Gill, professor in mathematical statistics at Utrecht university, gives this argument at a roundtable discussion:
“We should be collectively ashamed not to know anything about quantum mechanics. I would like to see all introductory texts in probability theory going a little into the physical (quantum) theory behind the geiger counter before using some data of alpha particle counts as an illustration of the Poisson process; I would like a discussion of the Bell inequalities together with a modicum of quantum mechanical background to show how elegant probabilistic reasoning shows that the quantum world is truly random (unless you would like to go for an even more weird non-local deterministic theory).” (1997b)
Indeed, also economists are familiar with the concept of Brownian movement, or the random walk, and use this model for example in analysis of the stock markets. Or in the labour market, with labour supply LS and employment LE, unemployment is u = 1 - LE/LS: but u then basically is a probability, since the model does not provide an additional explanation why one person works and the other doesn’t.
But Gill’s argument does not convince me. The point is: you may pose that nature would be such, but you don’t know for sure. You are still using only a model. The scientific challenge remains to develop a model that increases accuracy.
Yes, there is the Heisenberg uncertainty model that if you measure position then you no longer know speed, and if you measure speed then you no longer know position: and this model nicely captures a basic notion of uncertainty. But, try for a better model then - and take some thousands years more to do so. [64] [65] [66]
As a corollary, we can take a position on path-dependency (hysteresis) and chaos.
Some authors use the word ‘chaos’ in the sense of path-dependency. For example, a small variation in first conditions (starting point, parameter) can cause a widely different result - a butterfly flapping a wing can cause a tropical storm. Since we already have the term ‘path-dependency’ for this, we better reserve ‘chaos’ for the meaning of ‘seemingly random’. A chaotic system, in this proper sense, then gives a fully deterministic description, but the outward appearance that some variables would be random. Here it is strange that people who are in favor of ‘chaotic modeling’ also use this to be against determinism.
Path-dependent and chaotic models can be useful. The orbit of Earth around the sun looks solid, but over the billion years it seems pretty random. There is Schrödinger’s cat model that shows the macro world depending upon a micro state. There are the strange models in history and biology, where for example a meteor wipes out dinosaurs. OK, all these models exist, and they can be real good descriptions of true states of nature. But all this does not disprove the definitory deterministic predisposition of science. If you would run the movie again from the start (which is currently said to be a Big Bang, but I don’t know about that), then you would get, by the models that science tries to develop, the same result. If you would argue that anything else might pop up, and your mother could be a dinosaur with a pig’s head, and if you would develop models that would show this, then you are quite in danger of being out of science. (You would drop out on this definition, but could be in on the other criteria.)
Concluding this section, we find that definitions indeed guide our understanding of nature. The definition of science itself guides our perceptions - for example when it guides us into taking quantum mechanics as a model only instead of as ‘reality itself’.
A reason to be strict about this definition of science is that people, who would argue that nature is basically random, would also tend to reject deterministic results of science. A deterministic result of science is for example (1) that divergent indexation of tax exemption and the standard of living causes a tax void, and (2) that the existence of a tax void can be used to ‘abolish taxes’ without costs. It would be a pity if this result were to be rejected because of a fundamentalist ‘random view of the world’.
Our subject is the political economy of western welfare states, and in particular employment and inflation aspects. This subject is quite complex, and we must be modest about our results. Of course we can use statistics of the national accounts, and thus indirectly we use the statistical labour of thousands of statisticians, and indirectly the results of thousands of firms and of millions of citizens that filled in their tax forms. Economic literature provides a wealth of models and interpretations of these data. In my case, I also rely on my own experience in constructing a national economic model. All this, however, does not mean that we can forget about modesty, on the contrary. Nevertheless, it is my conjecture that we can achieve a more enduring result than just awareness of complexity.
What is interesting in economic discourse is the concept of ‘stylized fact’. When an economist observes some regularity, he is rather inclined to use that term. We shall use the term more conservatively, and we are hesitant about observing regularities. But we also can fruitfully employ the term when there is a regularity indeed. In some cases, when the regularity is so strong that our loss function comes in the epsilon zone, then we even can switch to definitions.
So we adopt the methodology:
(a) state what we consider to be the stylized facts
(b) define our concepts so that the stylized facts are covered by definitions
(c) develop theorems and proofs
(d) link back to conclusions about reality.
A proposition - as a statement on reality - can be regarded as a mathematical theorem about/within a model of stylized facts. When there is a tautology, we attain truth by definition.
We here deliberately refer to Bochenski (1956, 1970:20): “The word ‘proposition’ has been variously used, (...) nowadays commonly as the objective content of a meaningful sentence”.
Some students of the History of Economic Thought will see a clear resemblance of above methodology and what Schumpeter called the “Ricardian vice”. Quoted by Tintner (1968:7):
“His interest was in the clear-cut result of direct, practical significance. In order to get this he cut this general system to pieces, bundled up as large parts as possible, and put them in cold storage - so that as many things as possible could be frozen and “given”. He then piled one simplifying assumption upon another, until, having really settled everything by these assumptions, he was left with only a few aggregative variables between which, giving these assumptions, he set up simple, one-way relations so that, in the end, the desired results emerged almost as tautologies.”
This is almost exactly what we shall do, except that we generate tautologies.
Step (d) comes closest to the Popperian falsificationist criterion. Our deductions need not be insulated against testing, even though this present book abstains from econometric testing since we are too much involved in creating our concepts and constructing consistent and useful propositions. [67] Abolishing the Tax Void is a good and cheap test anyway for the relevance of this analysis.
It is useful to keep Solow’s comment in mind:
“There is something deeply satisfying - not to say suspicious - about any proposition that seems to deduce important assertions about the real world from abstract principles.” (1976:148)
So, advisedly, the reader better checks what we are doing here, and governments should run their own regressions and models before they make policy decisions. But of course I only dare to present my results here since I am confident that they, in the hands of competent and true scientists, allow a real advancement.
In his essay “A discipline not a science” (1983:365-375), John Hicks argues that economics is too far from the accuracy reached in the material sciences, and explains that he cannot ‘altogether’ deny that he himself has converged on a ‘critical’ attitude. This attitude concentrates on the clarification of terms, i.e. their definitions, also by using quite unrealistic models. For example: “Though the concepts of economics (most of the basic concepts) are taken from business practice, it is only when they have been clarified, and criticised, by theory, that they can be made into reliable means of communication.” (p372-3).
Hicks then concludes that economics is a Discipline. His quote of Keynes (in II.7) above is taken from these pages. My position on this is twofold - the position of hard science with soft data. On one hand I embrace the critical attitude. Indeed, we should develop sound definitions, and remain critical about how these are applied in communication. That is the meaning of the Definition & Reality methodology. And it brings us far, since we can advise to abolish the Tax Void without running regressions and a computer model. On the other hand, Tinbergen’s efforts have not been in vain, and models with estimated coefficients are useful tools for policy analysis. For example, some economists may reject the existence of a Phillipscurve, and all economists should be critical about the data and the parameter values, but such a relationship remains useful in a macromodel that is used for evaluation of policy alternatives. It would be curious to accept the concept of a ‘model’ and to accept other relationships like a consumption function, and reject the use of a Phillipscurve: even though the uncertainties are quite comparable.
In other words, our method remains econometrics, even though we end here with an increased awareness of the role of definitions. We are just in the phase that running regressions is useless if the model is no good. Regressions come in only when we have a good candidate, and regressions even might benefit from some definitory relationships. We even would like to do those regressions ourselves if we had the data and the time. So, for now, let us first develop what we conjecture to be the proper model.
There is the useful distinction between the structural and reduced form:
· the structural form represents actual relations as good as possible,
· the reduced form gives the simplest representation, with the interaction minimised.
With y a vector of endogenous variables, x a vector of exogenous variables, and f and g functions, then a structural form is y = f(y, x) and a reduced form is y = g(x).
Since econometrics can only approximate reality, the true structural form can only be approximated. What we consider to be a structural form is an intersubjective consensus. We anyhow have to adopt an approximation, which means that many factors have been removed. However, for two models we can often clearly see that one is simpler than the other, and then we can usefully apply this distinction between the structural and reduced form.
The distinction between structural and reduced form also affects the structure of this book. The next chapters concern the structural form, actually starting with the textbook IS-LM model. We relax the assumption of homogeneous labour, and introduce heterogeneous labour. First we look at labour supply only. Then we look at supply and demand, and at the equilibrating dynamics, which causes the topic of the Phillipscurve. We show how the Phillipscurve and the Constant-Wage-Inflation Rate of Unemployment (CWIRU, a.k.a. NAIRU or natural rate) shift as a consequence of minimum wages or poverty. We then relate minimum wages and poverty to developments in taxation. The co-ordination failure on taxes and minimum wages not only causes the internal imbalance on the labour market, but also an external imbalance, with international trade.
The discussion of the structural form results into the need for more scientific clarity. Though much seems to depend upon empirical parameters, some aspects however are more fundamental. This leads to the discussion of the reduced form. We first develop a theorem on the influence of taxation on employment and unemployment regimes in welfare states. Since taxation depends upon social choice, we then discuss Arrow’s theorem on social choice (structural form again). We also note that there may be a confusion about inefficiency and the existence of a ‘free lunch’. Having established the possibility of rational social choice, we then develop a theorem on stagnation in the policy making process (reduced form again).
In chapter 8 we stated: “If the government on the one hand would desire to use the results of scientific advice for its budget process, and on the other hand would not opt for an Economic Supreme Court, then its definitions would be logically inconsistent, and it would thereby tend to create a cause for dishonesty and improper manoeuvreing and thereby corrupt its processes.”
We can directly apply our Definition & Reality methodology. The point is that desiring for a scientific base and not making a Court is logically inconsistent. Parliament and President may ‘define’ their ‘Council of Economic Advisers’ as ‘scientific’ but when there are little safeguards, then reality takes over, and the Council will de facto not have sufficient power to resist political meddling.
The appendices contain an example draft for a Constitutional Amendment for an Economic Supreme Court and a description, taken from the White House internet site, of the CEA. The difference should be clear.
Law-givers know: If a law does not fit logic and reality, then people will see themselves forced to ‘break’ the law. “You are damned if you do, and damned if you don’t.” People in such situations will tend to grow dishonest, since it is often easier to massage events rather then clearly state that the law is impossible and go on strike or whatever. They don’t see it as ‘dishonest’, but as ‘flexible’. And once people are on that road, they will rationalise their behaviour by thinking that this is the way that the world works, and become more willing to perform other acts of dishonesty.
Conversely, once sufficient safeguards are in place, then the Council is de facto an Economic Supreme Court (even if it does not have that name). With a properly defined scientific base for the budgetary process, economists could also more confidently predict the economy’s course, since there would be less random noise and chaos about the application of known knowledge.
We consider all Western economies, or, more properly with Japan included, the OECD area. Hence, the student of this book will expect masses of OECD data, and masses of structural models of the OECD countries, or at least a model for the whole OECD area. There is none of that. We in fact use only some example data for the small country of The Netherlands. Why is that ? And how can we possibly utter our ambitious claims ? The answer to these questions is fourfold:
· there are mathematical theorems and proofs for the reduced form of a typical welfare state
· we use some key properties that will be documented here
· this chapter on methodology explains the validity of the method
· for the data and structural models we refer to ‘existing economics’.
The approach of this book is to use logic in order to circumvent the uncertainty of parameter estimates. Though the book doesn’t give full statistics, it is conjectured that the theorems capture the stylized facts. A proposition - as a statement on reality - can be regarded as a mathematical theorem about/within a model of stylized facts. When there is a tautology, we attain truth by definition.
Our first proposition establishes conditions under which both unemployment and full employment are possible. This relates to the partial arguments of economists about the labour market. Our second proposition gives the integral argument, or general theory, how (un-) employment situations are managed. The employment regime can be chosen by conscious choice, or there is lack of knowledge. Lack of knowledge forks into two cases. With full employment, the situation is dubbed ‘chance’. With unemployment, it is called a co-ordination failure.
It is useful to state that our point of departure was not mathematical economics itself. This book has been written against the backdrop of the voluminous studies Central Planning Bureau (1992a&b) and Colignatus (1992). It is from this experience that these two propositions have been selected as being of foremost importance. We want to focus on main mechanisms that block full employment and prosperous growth in modern welfare states. It is thought that the two propositions, in a sense simple but in another sense complex, help to clarify a fruitful direction for both analysis and policy improvement.
To be sure: this approach does not imply a rejection of time series econometrics ! I am an econometrician myself. Below I will e.g. develop a definition of ‘risk’ that deals with uncertainties - and in my view the 95% confidence interval should be replaced by an interval based on a well specified loss function. So I am supportive of uncertainty approaches. However, econometric models also contain definitions and institutional equations, and it is my conjecture that these have not gotten the attention required. In particular the regime switch of 1950-1970 to 1970-2005 will be difficult to determine by time series methods. Studying marginal changes within a regime will not uncover results about the switch. It would be wrong if time series analysts would only accept time series as data, and not such regime states. The Definition & Reality methodology then can help us out. [68]
Governments that become interested in the present analysis will no doubt require that it is tested against the data of their own country. This is advisable indeed. However, the claims of this book are primarily mathematical certainties, and additional empirical data will mainly provide didactic assurance. Since country parameters are different, practical policy must rely on the structural models of course, and data will be needed for detail decisions. But at an abstract level, the developments would be similar.