Evaluating Alternative Techniques for Forecasting Industrial and Occupational Employment
This paper offers three different regional output-by-industry forecasting techniques (time series, Social Accounting Matrix (SAM)-based, and Computable General Equilibrium (CGE)-based) and two different occupation-by-industry matrices (national and state geographies) for use in the creation of indus...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en |
| Published: |
LSU
2014
|
| Subjects: | |
| Online Access: | http://etd.lsu.edu/docs/available/etd-04092014-110652/ |
| Summary: | This paper offers three different regional output-by-industry forecasting techniques (time series, Social Accounting Matrix (SAM)-based, and Computable General Equilibrium (CGE)-based) and two different occupation-by-industry matrices (national and state geographies) for use in the creation of industry/occupation employment forecasts. Estimates are compared to actual data from eight years from 2001 to 2010. OLS regressions are run to determine how well modeled employment estimates fit actual employment for the state of Louisiana. A meta-analysis style regression of the R-sqaured values on model characteristics (accounted for using Boolean dummy-variables) determines that industrial output forecasting techniques do not provide statistically different R-squared values, but that models which use the state level occupation-by-industry matrix constructed for this paper should expect a statistically higher (by about 3.5%) R-squared value. |
|---|