A study of the feasibility of using forecasting models to characterize the supply of contract construction labor

The purposes of this study were (1) to develop comprehensive construction labor supply forecasting modelse for several SMSAs in Texas and (2) to determine whether such models could be used to characterize construction labor supply in these same SMSAs. This was undertaken to meet growing needs in business and government for increased quality in manpower projections, this leading to a more efficient and effective utilization of manpower resources. The first phase of this study was to develop comprehensive models suitable for forecasting purposes. Criteria were established for the acceptance of a model as forecasting satisfactorily. This phase also involved the collection of the data and seasonally adjusting it. using the ratio-to-moving-average technique. The models were developed through regression analysis techniques utilizing the stepwise procedure. The procedure enabled the user to control the significance level of variables remaining in the final regression equations. A total of sixteen such forecasting models were developed that met the established criteria. Of these six were single SMSA models and ten were multi-SMSAs models. The second phase of this study involved analyses to determine differences, if any, In the characteristics of SIMAs that were combined into multi-SMSAs models. This was accomplished through comparing the sum of the ESS of full models of the SMSA in a multi-SMSAs model with the difference between this sum and the ESS of the reduced model. An F test analysis was used to Identify the remaining differences between the SMSAs involved in multi-SMSAs models. This study attained its stated purposes. It has proven that it is possible to aggregate some SMSAs into a single forecasting model. The feasibility of using forecasting models to characterize some economic and other differences among the SMSA construction labor supply through detailed model analysis has been demonstrated for all of the seven SMSAs studied.