
27 February 1998 Estimating Shipyard Employment Levels Using Structural Modeling By Midn 1/c Titus Fortner
Current models for predicting the total employment of naval shipyards are inadequate. An econometric modelling approach is used by NAVSEA, which utilizes the ship contracts and start dates as inputs and forecasts the predicted employment as its output. The problem with their model is that it allows for more variation in the workforce than actually exists in the real world. This means that the results aren’t as credible as they need to be. There are several potential causes, the most likely being that there is a cost to hiring and firing of employees that causes the shipyards to hang on to more people than they can presently employ with the expectation of future contracts. Another potential cause is the current model does not take into account the seasonal fluctuations of work that occurs. The third cause is the current model fails to account for is the fungibility of workers in the same occupation among several contracts. Another approach to this problem is to use a structural model like a linear program. This model reflects the management structure of a shipyard in employment of the workforce. The variables to be used are: hiring employees, firing employees and working employees overtime. By minimizing the costs associated with each of these variables over several employee types, several contracts and many time periods, an optimal solution to the model is computed. The paper describes the structure of the model and how the coefficients were arrived at for the model. It compares the behaviour of the structural model with an econometric model and with the observed real world data. 
