An Urban-Regional Employment Demand Model

Michael W. Babcock
1974 The Review of Regional Studies  
In his article "An Approach to Metropolitan Employment Forecasting" Rich ardson pointed out the lag in the development of subnational area economic fore casting behind other branches of economic forecasting.^ This seems paradoxical in view of the need for such models in formulating public policy. The regional im pact of federal expenditure programs, the formulation of local economic policy, and the adoption of rational education policies represent problem areas which need a local economic
more » ... ocal economic forecasting system as an information input. In the article cited above, Richardson discussed the limitations of the exist ing local forecasting techniques, so I will only briefly review them here. The existing techniques include (1) naive models, (2) export base models; and (3) in put-output models. The naive models, usually involve trend extrapolation of some variable such as regional employment, income, or output. These models may re sult in reasonably accurate forecasts in the short run, but yield very poor long run forecasts since they blindly project the past. The major weakness of export base models for prediction purposes is the dynamic instability of the base-service ratio. The fundamental drawbacks of input-output models are the well known empirical implementation obstacles. These include obtaining the regional inter industry coefficients matrix, which implies that the researcher must generate his own data by some type of survey method. Non-survey techniques designed to simulate regional interindustry coefficients have so far met with only modest suc cess.^ Another difficult problem is the development of techniques to forecast re gional final demand. The local emplojnnent demand model presented in this paper was developed at the Center of Advanced Computation, University of Illinois as part of a man power and educational forecasting system known as STEP I. The industry em ployment demand model served as a basis for detailed occupational demand fore casts for the state of Illinois.^ Thus the model presented here is a local industry employment demand model. As Richardson has pointed out, the output of such a model, a vector of forecast industry employment, can serve as an important first step in forecasting other economic information such as occupational demand, in come distribution, and net migration.^ The results of the Illinois study are present ed for illustrative purposes, but the application of the model to metropolitan areas wiU be discussed. THE MODEL In the development of the theory of regional growth, demand has received far more attention than supply. The existing subnational area employment fore-*Assistant Professor,
doi:10.52324/001c.10427 fatcat:vqxaspzkhfhltoa3ua3ri7e7mu