Showing posts with label econometrics. Show all posts
Showing posts with label econometrics. Show all posts

17 April 2010

Wealth Dispersion and General Thoughts on the Future of Economics on a Saturday Afternoon


Here is an interesting graph of wealth distribution, or dispersion, as I call it from Cherchez La Verite.

I am not sure I agree with his conclusions or even his premise, not because I disagree but because it requires some thinking and leisure to digest it. But the data is most interesting.

I wonder if any of the quant economists have performed simulations on virtual populations, and then examined the results of varying different tax rates, and concentrations of wealth because of fiscal policy and regulatory structure, among other things.

I have an hypothesis that great concentrations of wealth lead to economic stagnation, but I am afraid that I have not the means or the talent anymore to conduct that type of research.

The difficulty in a study like this is that the assumptions are greatly magnified into the results. If you assume certain buying, spending, and savings behaviours, the downstream impact can greatly alter, and even distort, the outcomes.

And when people reason through this verbally, rather than perform a structured simulation based on transactions, the distortions increase by an order of magnitude or more based on their own biases.

I used to create simulations like this all the time, for industrial and commercial purposes, and also did a decent amount of econometric modeling. So I am sure someone is doing it somewhere. But I suspect they are doing it in think tanks and places where the outcome is predetermined by the basis of their grant.

Concentrated wealth magnifies the needs and predispositions of the holder. Since the amount they require for basic necessities can only consume so much, one would think that the amount spend on the aggregate of necessities will eventually be reduced. And what they do with their excess of necessity wealth is going to be greatly influenced by their character. Are they a gambler, who inherited the wealth? Are they productive and beneficent? Are they dissolute and venal?

And what about government? Taxation can concentrate enormous wealth in the government. What sort of government does one have, or does one assume? Are they warlike, productive, redistributive, and how corrupt? What about corporations? They can be like small governments, and levy taxes through monopoly and persistent frauds. How are they managed? Corporations are not rational machines, as the efficient market hypothesis would probably presume. Indeed, corporations are often much worse than governments in terms of sheer blockheadedness, greed, and short-termism.

Hard to say. But there is a related field of study in decision making theory, which looks not at wealth but the distribution of decision making power in organizations. It is concerned with the validity and effectiveness of decisions made across a range of broader consensus to a narrow oligopoly and even a great man dictatorship.

The general observation I came to in this study was that decisions tend to be more valid depending on the quality of the information, the facility of the evaluation of it, or intelligence/learning/experience, less the biases and distortions.

A decision becomes a little better if the information is more widely dispersed and a variety of actors can exchange freely in increasing and refining it. There is a point of decision dispersion where the returns not only diminish, but become counterproductive because of the noise and inability of new actors to add value, and actually detract from the process. But finally what I found interesting is that in the aggregate personal error, bias and distortions tends to diminish quickly as a detractor from the result, assuming a non-homogeneous population with some independence of thought.

So too this same sort of study can be applied to the concentration of wealth, since wealth is power. But it is even more interesting because spending habits will vary since the percentage of spending on essentials changes much more slowly than wealth can increase.

And how one assesses the outcomes is also essential. What is thought to be a 'good outcome?' Not necessarily in a rough measure like aggregate GDP, but perhaps GDP with modifiers like the median wage, and a poverty level of essential spending. This is important because so often economic policy arguments are presented with the goal of optimizing short term GDP.

Alas, I have little hope that this will be done now, for the US has had a leadership role in quantitative economic studies, and their work has been twisted generally into the service of whores, robber barons, and gamblers as the speculative society reaches a crescendo. But some day this too will change.