1998:The year in review: Which questions are worth asking and answering?

[Note for bibliographic reference: Melberg, Hans O. (1999), 1998: The year in review Which questions are worth asking and answering? , www.oocities.org/hmelberg/papers/990106.htm]

1998: The year in review
Which questions are worth asking and answering?

by Hans O. Melberg

Introduction
The beginning of a new year is a natural time for reflection on the past academic year. The basic objective is simply to "learn from past mistakes." Besides, by looking at the 20 papers written in 1998 as a whole, I might see some patterns that I was unable to see when working on each paper in isolation. That is, general themes and questions that may unite the papers. This reflection on the general underlying questions may then form the basis for asking whether the issues I write about are worth studying. In short: Which questions are worth asking?

The papers
As mentioned I wrote about 20 papers in 1998, compared to 37 in 1997. Among these there were 2 longer articles ("Explaining the appeal of protectionism" and "Is ethnic conflict the outcome of individually rational actions?"), 7 reviews and 11 observations. The list in reverse chronological order is as follows:
• Decision making and infinite regress: More complicated than I thought is was!
• Rent seeking, social waste and economic models
• Definitions of economics: A short and uncritical introduction
• Explanation, benchmarks, Bayes rule and logical impossibility: Notes related to economic fluctuations
• Psychology and economic fluctuations: Pigou, Mill and Keynes
• Political culture as an explanatory variable - A (very) brief historical overview
• Is it logically possible to make a rational decision? Optimal collection of information and "as-if" arguments
• Not bad, but more popular than it deserves: A Review of Kindleberger's "Manias, Panics and Crashes: A History of Financial Crises"
• Entertaining and suggestive about the business cycle - Review of Lars Tvede: Business Cycles: From John Law to Chaos theory
• Arguments searching for proofs: A Review of Galbraith's A Short History of Financial Euphoria
• Explaining the appeal of protectionism: From mercantilism to today's popular opinion
• Searching for the surprising and reliable truths: A review of Schelling
• Is ethnic conflict the outcome of individually rational actions?
• Twisting the concept of rationality to fit collective action in ethnic conflicts
• Principal-agent theory: What happens when we have lazy, risk-averse and dishonest people who know more than us?
• Cooperatives: A short model with surprising implications
• Pay according to need or effort? About models in economics
• Visual presentation of non-linear correlation in n-dimensions: A speculation
• Why (not) use the cultural approach? Review of Tucker, Part II
• Why (not) use the cultural approach? Review of Tucker, Part I

Number of hits
The pages receive about 430 hits every week (slightly down from 450 last year). Some papers, of course, were more popular than others were. The list of the fifteen most popular papers is as follows:

The fifteen most popular papers in 1998 (out of the 95 papers available)
1. Explaining the appeal of protectionism: From mercantilism to today's popular opinion (697 hits in 1998)
2. A review of reviews: Making Sense of Marx (685)
3. Lessons from History: The collapse of the Soviet Union and the end of the Cold War (658)
4. Three arguments about rational choice theory in sociology (468)
5. Three statistical examples: Boeing, sex and education (388)
6. Inflation: An overview of theories and solutions (358)
7. Is ethnic conflict the outcome of individually rational actions? (358)
8. Organic explanations: Review of Brzezinski: The Birth and Decay of Communism (355)
9. Dynamic counterfactual correlations: Comments on Caplan's "The Mensheviks' Critique of Bolshevism ..." (355)
10. Unemployment: Micro- or Macro-theories? (333)
11. Russians about Russians: A sample of views on cultural traits (312)
12. Arguments searching for proofs: Review of Galbraith's A Short History of Financial Euphoria (312)
13. The Cultural Approach to Russian Politics: How Reliable? (291)
14. Sovietology: Why the totalitarian model is the best (272)
15. Evolution or Creationism: Does science and religion compete in the same arena? (263)

Reflecting on the information
Since the main aim of the page is to develop my own interests, I am not too concerned about the number of hits and -- like last year -- I have done little to promote the papers. I am, like I was last year, more concerned about the papers that receive most hits. First of all, many of the most popular papers were written long time ago when was an undergraduate. This applies to the papers on inflation, unemployment and rational choice theory in sociology. These papers are not my best papers. I have thought about removing them, but I decided that a "warning" on the top of each paper was enough to prevent people from taking them too seriously. Second, some of the more recently written popular papers, are simply not very good, for instance Three statistical examples: Boeing, sex and education.

I am, however, more pleased to see that some of what I consider to be acceptable papers also are on the most popular list. This includes the most popular paper in 1998 Explaining the appeal of protectionism: From mercantilism to today's popular opinion and the seventh most popular Is ethnic conflict the outcome of individually rational actions?. What I consider to be the best researched paper, however, only came thirteenth (The Cultural Approach to Russian Politics: How Reliable?). And, as always, I was disappointed to find that some relatively decent papers did not make the list. For instance, The information required for perfect prediction was number 45 with 141 hits and Against correlation was only number 55 (i.e. average) with 113 hits. The explanation, I guess, is relatively simply. First of all, the topics appear not to interest the general reader (although they should!) and the titles -- at least Against Correlation -- may deter some readers.

The themes
The papers from 1998 may be divided into the following thematic groups:
- 5 papers on economic fluctuations
- 4 papers presenting formal models in economics as a preparation for a larger paper on formalism in economics.
- 3 papers on economics in general (definitions and so on)
- 2 papers on the problem of infinite regress in decision-making
- 2 papers on the use of political culture to explain history
- 2 papers on ethnic violence (and rational choice theory)
- 1 "various" (a speculation about the visual presentation of multi-dimensional information)
Let me say something very briefly about these categories and papers.

Economics fluctuations 5
• Explanation, benchmarks, Bayes rule and logical impossibility: Notes related to economic fluctuations
• Psychology and economic fluctuations: Pigou, Mill and Keynes
• More popular than it deserves: A Review of Kindleberger's "Manias, Panics and Crashes: A History of Financial Crises"
• Entertaining and suggestive about the business cycle - Review of Lars Tvede: Business Cycles: From John Law to Chaos theory
• Arguments searching for proofs: A Review of Galbraith's A Short History of Financial Euphoria

This was a series of book-reviews in preparation for a larger paper that I have decided not to write since I found the topic too demanding to be answered usefully. Initially, I was interested in whether economic fluctuations should be explained as a rational outcome of individuals operating within a free market economic system, or as the outcome of (non-rational) psychological mechanisms that influence economic confidence and hence economic activity.

Formalism in economics: Models 4
• Rent seeking, social waste and economic models
• Principal-agent theory: What happens when we have lazy, risk-averse and dishonest people who know more than us?
• Cooperatives: A short model with surprising implications
• Pay according to need or effort? About models in economics

This is a topic on which I will continue to work on. I am gradually becoming convinced that "non-fat" modeling is useful, but I remain unsure about the degree to which economics has become a mathematical subject. One problem, which emerge in several of the papers below, is that the supposed advantage of these models (being able to derive surprising implications from the assumption of rational behaviour), turns out to be less interesting when these implications are not true empirically. Maybe labour cooperatives should reduce their production and employment when prices increase, but empirically speaking they do not do so. The model on rent seeking yields the surprising result that a firm with persuasive arguments needs to spend the same amount of money as the firm with not so persuasive arguments. Yet, how interesting is this when we do not observe this empirically speaking?

My answer, based on working within these models so far, is that the use of models is not mainly to derive policy-implications or true explanations, but as a frame for discovering missing links and variables in the explanations. Formal modeling enables us to specify in ever-growing detail the conditions that have to be true for the argument to be valid. This, I think, is a valuable exercise.

Economics, General 3
• Definitions of economics: A short and uncritical introduction
• Explaining the appeal of protectionism: From mercantilism to today's popular opinion
• Searching for surprising and reliable truths: A review of Schelling

"Definitions of Economics" should be read together with "What is economics" from 1997. "Explaining the appeal of protectionism" is longer, but not much better, than Why don't people believe that free trade is good?. In fact, I believe there is a small mistake in the essay (where I mistakenly distinguish between specie and gold). There is also a questionable point in the review of Schelling. I was slightly too quick to accept his argument about the ski-lift and I changed the example in a way that may distort his point. I have discussed this issue with several people and I am still unsure ...

Rationality, Decision-making, infinite regress 2
• Decision making and infinite regress: More complicated than I thought is was!
• Is it logically possible to make a rational decision? Optimal collection of information and "as-if" arguments

This topic grew out of an interest in justifying the use of psychology in economics. The idea was that sometimes it was logically impossible to make a rational decision, so the decision that was made cannot be explained as the outcome of a rational choice. Alternative theories of behaviour were needed, such as psychological mechanisms. Yet, as I explored this argument I first became less sure that the common argument was sound. Maybe the rational collection of information is possible? However, more research revealed that there were several other infinite regress problems that also created problem for the possibilities of making a rational choice. For instance, I believe the formation of beliefs for a given set of information may run into some kind of infinite regress problem. (If I do X, then he does Y, so I should do Z, but he knows this so I will not do Z in order to fool him, but he may know that I think like this so he will not be fooled, but if he knows that I know that he knows this then it would not be rational for me to do Z and so on ... Or: I am rational and I know that he knows that I am rational. He also knows that I know this, and I know that he knows this and he knows ... and so on in an endless regression).

Political Culture 2
• Political culture as an explanatory variable - A (very) brief historical overview
• Why (not) use the cultural approach? Review of Tucker, Part II
• Why (not) use the cultural approach? Review of Tucker, Part I

Most of my writings on this topic were in 1996 and 1997, but in 1990 I will once again focus on this topic with the view of writing a Ph.D on the arguments. I have recently finished a large article in Norwegian based on my 1996 paper.

Ethnic conflict and violence, nationalism 2
• Is ethnic conflict the outcome of individually rational actions?
• Twisting the concept of rationality to fit collective action in ethnic conflicts (Review of Hardin)
Most of the criticism of Hardin, I believe, is important and valid. I may have given too little attention to his valuable contributions (on the relationship between norms and rationality), but I felt the book was contradictory and generally not very good. More seriously, the recent growth of rational choice explanations of ethnic violence is flawed by several mistakes. There is an insufficient distinction between what is rational for the group and what is rational for the individual. There is also an overemphasis on the factors that confirm the rational explanation, and a relative neglect of serious alternative. Often the contrast is with a straw man "ancient hatered" model that is not really presented in any depth. Finally, the empirical evidence is not used properly to "test" the theory. Often only confirming evidence is selected and the empirical evidence showing that the links between variables is much weaker than the theory implies, is ignored. In the second half of 1999 I will write a paper on this and try to get it published.

My own papers are, however, far from perfect. In the large paper (Is ethnic violence ...), I have doubts about the relevance of the KMRW argument, and in the appendix I suddenly change from working with a "tit-for-tat" strategy to a "trigger" strategy. The reason, I believe, was that I have read the proof using both strategies. Moreover, appendix III is not proof-read and the whole essay was written under strong time-pressure.

Overall: What is the question?
For some reason many of my papers end up discussing rationality, methodology and the nature of explanation. On the issue of economic fluctuations I set out to contrast rational and non-rational explanations. The same applies to ethnic conflict: Is it the outcome of rational or irrational forces? On political culture I try to examine the reliability of one alternative to rational choice explanations (i.e. cultural explanations). As for the nature of explanations, it seems like the major topic is whether it is best to work within a general (equilibrium) frame or whether we should focus on mechanisms. The issues of parsimony and reliability are also relevant and discussed in some of my essays on economics.

Is it worth spending so much time trying to determine whether something should be labeled the outcome of rational or irrational forces? I am unsure. There are at least two questions. First, whether it is a question of mere labeling. Second, whether the debate is one of distinguishing between different causal forces.

The second is surely worth discussin. What are the causes of economic fluctuations? What are the causes of ethnic violence? The answer to these questions are important both from an academic and a policy point of view. But, the first may seems more questionable. For instance, my criticism of Hardin is mostly a matter of labeling: I do not want to use the label rational for an action that grows out of irrational collection and use of information. This may be a problem of words, yet I tend to think that labeling also matters.

Another point is this: "What is most important in an explanation: The motivation behind actions or the context within the decision was made?" Of course, both are part of an explanation, but as I became increasingly aware of as I studied economic fluctuations, sometimes the structure of the situation is such that the question of motivation becomes less important. For instance, small deviations from rationality can in principle create large economic fluctuations (given nominal and real price and wage stickiness, see D. Romer's book Advanced Macroeconomics). Thus, the key need not be exactly what motivates us to act the way we do, but how the market is structured. We need to consider not only a collection of individuals, but that the activities are structured in a market. There are factories, distributors, speculators and customers. The links between these are important to explain economic fluctuations, not just individual motivations. For instance, Schelling describes how small changes in one level (say consumption) can lead to a large change on another (say investment by factories). [Herbert A. Simon also discusses this issue in his book "The sciences of the Artificial" in the essay on economic rationality].

When I wrote "the key" in the paragraph above, I was actually making a comment on the nature of an explanation. For some time I have worked using Jon Elster intuitive definition of an explanation as the answer to why something happened. And, since I also have adopted his methodological individualism (only individuals make things happen), I gravitated towards the question of what motivates individuals. Although I still work within this tradition, I now believe that motivation is not everything there is to an explanation. Try, for instance, the following very informal definition of an explanation: "An explanations is a statement that makes you say 'Oh, now I understand why it happened.'" . Say a rich person has been found murdered. The general motivation for the murder may be relatively obvious (greed), but you do not think this "explains" the murder, especially since (assume) the person in the will has a good alibi. Now, if somebody then tells you that another person (mistakenly) believed that he was in the will, you may exclaim "That explains it!" The general motive was always clear, but we did not believe that we had an explanation until we knew the beliefs of the people involved. I do not claim to have discovered anything significant in this story, only to say that personally I might have concentrated too much on the issue of motivation and not enough on the other variables that constitute an explanation.

Did I do what I said I would do?
In my review of 1997 I promised to write at least two articles (Review of 1997). I have written two in English and one in Norwegian so I delivered on that promise. I did not, however, manage to read all the "classics" that I wanted to read. The book "Hume and the problem of Causation" is still only half-read. M. Weber's books are still largely unread, as is Tocqueville's "Democracy in America". Scarcity of time is only half the explanation. It is a question of giving these books priority over more urgent (but less "important" readings). Maybe this should be my promise for 1999?

The next year
In the first half of 1999 I shall write several essays dealing with the problem of endless regression in decision-making. I then want to relate this problem to both economic fluctuations and ethnic violence. The specific question with respect to economic fluctuations is whether the problem of endless regression leads to psychological theories, which in turn help explain fluctuations. For instance, the problem of costly information collection implies that imitations may be profitable - and this justifies herd-based explanations of fluctuations. Another example, is the problem of how far back in the chain we want to go (if he does X, I should do Y, but he knows this so he will do Z and so on - just like in a game of chess). Maybe the payoff from going backwards is relatively low since it has to be very precise (going two steps more than the other person may ruin your calculations i.e. you need to know exactly how far back he goes) that we act as if we were in a parametric environment. This creates over-investment cycles (Cob-Web models) like J.S. Mill and others predicted.

Endless regression is also relevant to the study of ethnic violence since several authors (for instance D. Laitin and B. Posen) claim that beliefs about beliefs (about beliefs and so on) are important to explain why some societies live in peace and others do not. The formation of beliefs is crucial and Bayesian updating may be important (see Weingast 1999). Yet, if there is an infinite regression, rational choice explanations are inherently limited, and it is these limitations that may be important to locate the mechanisms that create violence (or fluctuations).

In the second half of 1999 I will focus more on explanations of history based on political culture (as it has been used in Russia).

Conclusion
1998 was not a perfect academic year. My work and other matters ate most of my time. Moreover, I did not even finish one article that I believe was good enough to be published in a journal. This should change in 1999, and the first article being one on political culture in Russia. The second should be on the problems of rational choice explanations of ethnic violence. The papers published on this page, hence, will be small observations and reviews that could be labeled "thinking aloud in preparation for larger articles."

[Note for bibliographic reference: Melberg, Hans O. (1999), The year in review Which questions are worth asking and answering? , www.oocities.org/hmelberg/papers/990106.htm]