Recruitment Early Warning System. Phase II. Executive Summary.
Abstract
The purpose of this project is to provide OSD and the Services with a source of timely, reliable information on the near-term status of the recruiting market. Phase I determined the feasibility of developing and implementing an early warning system for military recruitment; Phase II research has yielded a prototype Recruitment EWS, automated on a microcomputer, which produces monthly forecasts of labor market conditions and enlistments and generates monthly assessment reports. Enlistment forecasting models have been developed for each service. The forecasting accuracy of the enlistment models is reduced when a new program or policy changes relationships in the market structure. Research on alternative methods of remediating forecasting errors caused by such 'regime change' indicates that it is possible to regain, within three to six months, the level of forecasting accuracy which existed prior to the regime change. An unemployment forecasting model has been developed which is the function of 15 leading economic indicators. It is a regression model with an ARMA error structure, and is estimated with national monthly data from May 1972 to April 1984. In out-of-sample forecasting tests the model quite successfully predicts imminent turning points several months in advance of occurence. The model's forecasts are relatively accurate, and are used by the EWS as a variable in forecasting enlistments.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 1985
- Accession Number
- ADA166569
Entities
People
- Christine Kennicott
- Peter Greenston
- Sherry Andrews
- Sigurd Hermansen