Consumer Procedural Economics

 

I am currently writing a book on a procedural consumer. This book expands on the following papers and is directly related to Informational Society:

I Published papers

A. Two-Stage Budgeting: A difficult problem. This paper has now been published in the Computational Economics. See below for details and if you wish click here for pdf version.

This paper has two main points:

1. People make their purchases item-by-item and not by bundles.

Consider grocery store with 30 categories and 10 items/categories
Example: Types of cereal boxes, types of fruit, milk products, and so on
Suppose for the sake of argument that your shopping list had 30 items -- one from each category.

a. Conventional grocery store (Item-by-item): You take a shopping cart and proceed through the isles stopping at each category and select one item from each category.
b. Now consider an alternative arrangement(Bundle): The grocery story arranges a column of shopping carts each with a different combination of the items in the 30 categories. The consumer proceeds down the column and selects the shopping cart with the bundle he desires.

What is the difference

a. Item-by-item: The number of comparisons the consumer must make is 10+ 10 + ... + 10 = 30 x 10 = 300
b. Bundle: The number of comparisons the consumer must make is 10 x 10 x ... x 10 = 1030. (An unbelievable large number)

2. Budgeting is an incremental adjustment process

Allocating money into categories is a quadratic process. For example, If you wish to allocate $100 into three categories a person must conpare the utility of (100,0,0), (99, 0,1), (99,1,0) and so on.  Faced with a large number of alternatives, humans use sublinear procedures.

Many successful student budgetors do not use spreadsheets or other writen records.  What they do is monitor the flow of funds in their accounts and make incrementat adjustments.  The two most variable categories and food and entertainment.  Some student have a feast of famine budgeting style where they feast for the the first couple of weeks after receiving funds and then sharply cut back (famine) the last two weeks.  An important aspect of budgeting is learning to live within one's means.

B. An Ordering Experiment. This paper has now been published in the Journal of Economic Behavior and Organization. See below for details and if you wish click here for pdf version.

To explain ordering I have prepared a short video: Sort

The bucket sort is a linear process in that the number of operations is a linear function of the number of objects to be ordered.

The most efficient binary comparison sorts require n log2 n operations to order n objects.
Suppose you wanted to sort 100 items and  both a utility calculation and a binary comparison took 1 second.  The time to do a bucket sort is 100 seconds and the time to do the most efficient binary comparison sort is 664 seconds.

C. On the Complexity of Consumer Decision Rules This paper has now been published in Computational Economics. For pdf version, click here

Human take seconds to execute decision rules not nanoseconds.  This means that using a rule such as a binary comparison to find a preferred item faced with a large number of alternative is prohibitively time consuming.  Consider a binary comparison operator that takes two seconds to operate and a consumer faced with 10,000 alternatives.  The time to find the optimal is 20,000 seconds, which equals 5.56 hours.  It is much more efficient for humans to initiate searches with set operators that require only one observation per set.  Markets are organized to facilitate set operators.  For example, department stores in physical space organize goods by categories and car lots organize cars by model.

Humans are also resonable efficient at shifting between set and individual operators.

4. Can consumer software selection code for digital cameras improve consumer performance? This paper has now published in Computational Economics. For pdf version, click here

Information: Economics has a concept called perfect information that is incompatable with human information processing capabilities.  The limit of data that could be given to a human for processing to make a decision is the Heisenberg uncertainty principle.  Clearly the amount of data would total overwhelm human processing capabilities.  Consumers process a miniscule fraction of the data with which they could be provided.  Indeed, consumers generally process a small fraction of the data available to them in the marketplace.  A better concept that perfect information is the information value of data that depends on
    a. The processing cost including the cost of acquiring knowledge ot interpret the data
    b. the ability of the data to discriminate among alternatives
    c. The reliability of the data

Creation of search code software is in its infancy.  The current generation is dominated by the idea that decisions should be weighted averages of alternatives.  This is correct for the end of the decision process, but not the beginning.  In the beginning the consumer needs to be educated about the product and can deal with sets and specific attributes needed for the indended use.  Our paper focuses on the initial part of a search

The codes for this paper can be viewed here

II Work in progress

4. Cyberspace Representations and Choice. This paper was presented at Digital Communities 20003 in Stockholm, Sweden in June 2003. For the revised pdf version, click here  I have given up on trying to publish this paper as it is far from standard economics.  Parts of it will appear in the book.

5. Paper for ESA Tucson 2006 conference, Repeated Price Search. We are now working on a revision, Spring 2008. Latest version is Decision Heuristics and Price Search. This paper attacks standard economic price search theory as an 'emporor without his clothes theory."  Through repeated searches consumer acquire data on relative prices that they use in future price searches.  The Internet is totally changing the nature of price search because at one site a consumer can view an order list of the prices of a large number of sellers.  Getting this paper published is like the others a real challenge.
6. Paper for ESA Tucson conference Oct 2007 : Sequential Consumer Theory An earlier version was presented at the  Behavioral and Experimental Conference in Lyon, France May 2007 We will again work on this topic 2008-2009.  The point of this paper is that researches should not ask whether consumers optimize or not, but how close do they come to optimization.  The ordinal nature or utility theory creates a fundamental measurement problem.