A Full Distribution Approach to Simulating the Industry Aggregate in the Input-Output Modeling Framework

A full distribution approach is presented as an alternative to conventional aggregate variable representation in input-output (IO) analysis, and its implications and viability are explored. The basic supposition is that those macrolevel variables that are simple or weighted averages of samples from microlevel component distributions are most often poor representations of their underlying populations. This is supported by a number of theoretical arguments, and an aggregation experiment employing the 1972 United States national commodity-industry accounts. A representation of the full distribution of microlevel components by a theoretical probability density function (pdf) is introduced as an alternative to the traditional point estimate, given the varied and strongly skewed observed production coefficient distributions.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Geography

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2014

Author dc:creator

• Jackson, Randall Wade

• Geography

Identifier

(UMI)AAI8324582

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/70641