• There are many people who find rigorous arguments more convincing. If you learn to construct rigorous-looking arguments you will be able to more effectively convince such people.
• Groups of people who share conventions of rigor have more effective ways of disagreeing and arriving at truth through discussion.

So, what is rigor? wikipedia:

An attempted short definition of intellectual rigour might be that no suspicion of double standard be allowed: uniform principles should be applied. This is a test of consistency, over cases, and to individuals or institutions (including the speaker, the speaker's country and so on).

This means that arguments should be explicit about when and where they are valid. In the opening, the mathematician, physicist and engineer arrived at different conclusions because they started from different premises. The engineer had assumed that the observed sheep was representative of all sheep in Scotland, and the physicist had assumed that each sheep can have only one color.

Explicit assumptions are useful because they allow the quality of your logic to be evaluated separately from the soundness of your premises / assumptions. It allows people to divide up the task of understanding the world, because then ideas can be connected to each other. For example, If I prove that "A implies B" and you prove that "B implies C", then if from experiments we find that "A is true", we can infer that C is true. If, however, we find that the "A implies B" proof is wrong, we can still hang on to the "B implies C" piece until we find some other way of proving or measuring "B is true".

Despite this great benefit, however, people do not usually make all assumptions explicit - there are simply too many assumptions to exhaust explicitly. This being the case, the communication of rigorous arguments requires an understanding exist between speaker and listener, in the form of a body of common assumptions. In the presence of such a convention, only assumptions new or contrary to the convention need to be stated explicitly.

---

A rigorous argument is composed of three types of statements - premises, logic and conclusions. When conclusions are wrong, either the premises or the logic is wrong. If conclusions are observed to be wrong and yet both logic and premises are correct, people get thrown into a panic. This is because it means that one of the unstated, implicit assumptions must be wrong. Erroneous implicit assumptions are troublesome because you don't know where else they could pop up.

The ideal rigorous argument only has explicit assumptions. This is an unattainable ideal, because in reality it is impossible to state all assumptions. The ways in which this shortcoming is overcome are complex indeed. wikipedia:

A variety of 20th century movements, including liberalism and neo-classicism, traced their intellectual heritage back to the Enlightenment, and away from the purported emotionalism of the 19th century. Geometric order, rigor and reductionism were seen as Enlightenment virtues. The modern movement points to reductionism and rationality as crucial aspects of Enlightenment thinking, of which it is the heir, as opposed to irrationality and emotionalism. In this view, the Enlightenment represents the basis for modern ideas of liberalism against superstition and intolerance.

Reductionism, liberalism and tolerance of diversity all help to manage the body of implicit assumptions. It is ultimately these attitudes that make the ideal of rigor practical as a guiding principle in reality.

Paper by Brad Delong:

international capital mobility over the past two decades (a) has expanded much more rapidly than almost anybody had predicted, but (b) has expanded in the wrong direction—the poor have not been borrowing from the rich to finance their investment and industrialization, instead the rich have on net been borrowing from the poor to finance their own consumption.

...

The big story is that the large net flow of capital from the rich to the poor countries of the world seeking high profits from reducing disequilibria between the wages and the relative productivity of labor has simply not happened. Instead, the principal thing that happened was an enormous flow of capital from the periphery to the poor...

...

The attraction is that the core—especially the United States—offers a form of protection for capital against unanticipated political disturbances. Since 1990 global investors have valued the American provided political risk insurance that they can obtain by placing their money in the United States more than U.S.-based companies have liked the idea of producing abroad in places where the wages of labor are lower.

I first saw this in Greenspan's book. He mentioned that poor countries tend to have higher savings rates, but did not pose political instability as a cause.

If you are Tanzania or Mali or even Bangladesh, it takes 4 percent of GDP devoted to national savings and domestic prices to produce a 1 percent real investment share of GDP when real investment is measured at standard international prices. This implies an extraordinarily tilting of relative price structures against the poor countries of this world: it requires enormous domestic savings efforts to get even tolerable amounts of real capital to use for development.

...

It also points out a defect in the thesis that one reason that poor countries are poor is that their citizens or their leaders or their governments have by and large chosen to consume rather than to save. That is simply not the case: savings rates on a national level have little or no partial correlation with prosperity. It is relative price structures, and thus real investment shares of GDP as measured in international dollars, that have this high correlation.

This I don't really understand. Do they mean that savings in developing countries have the effect of depressing the exchange rate, leading to a slower increase in real terms?

Especially relevant for my friends looking to work in Dubai and Abu Dhabi:

DeLong and Summers (1991) found that the post-World War II crosscountry dataset contained an extraordinarily strong correlation between growth and private investment in machinery and equipment. Public investment by state-owned monopolies did not do it. Investment in structures did not do it. The correlation was very strong in OECD-class and middle income economies.

HT: Marginal Revolution

Saw this cute fixture in a restaurant in NYC. Reminded me of a genre of hedge fund interview question - what is the relationship between the number of Y-connectors and light bulbs? Note that how the Y-connectors are attached to each other does not matter.