Forecasting Skilled Labor Demand in the US Construction Industry

Ensuring an adequate supply of skilled laborers to meet demand and avoid potential shortages or surpluses has been an issue of concern to construction industry professionals and researchers for quite some time. Poor industry image, declining wages, lack of training opportunities, and training time lag between new, unskilled laborers becoming skilled are a few of the many factors identified by researchers and professionals over the past 20 years as having contributed to skilled labor mismatch. Although construction unemployment reached record high levels in February of 2007 (27.1%) due to poor economic conditions in the United States, as we continue through the 2010's and into the 2020's, maintaining an adequate and competitive workforce that is able to meet the future skilled labor demands is an issue of importance to the construction industry.;Forecasts of skilled construction labor demand provide valuable information that can ensure that construction industry participants and stakeholders are aware of future labor force needs and are prepared to recruit, train, and retain an adequate pipeline of skilled laborers. Forecasts of skilled labor demand for the construction industry are important to ensuring a sustainable skilled labor workforce. Therefore, the main objective of this work was to make accurate and useful forecasts of future skilled construction labor demand.;Crucial to developing reliable forecasts is the collection of accurate and consistent data with which models can be developed and on which to base projections. Data were collected in this effort for five key independent variables (interest rate, material price, construction output, productivity, and real wage) and the dependent variable (labor demand) from a variety of existing data sources. The availability and quality of economic and construction industry data intended for use in the skilled labor demand forecast model is assessed and evaluated.;Productivity data, in particular, has historically been difficult to collect. For this effort, a new productivity metric was developed using labor and cost information from a sample of typical construction activities in the RS Means Building Construction Cost Data manual. The newly developed metric was used as input to the developed forecast model.;The model developed in this research used vector autoregression (VAR). VAR modeling was selected because of its ability to analyze multivariate time series data. The forecast model was successfully validated against two years of actual data.;Potential data trends for each independent model variable were developed. Various combinations of the potential trends were used in the model to formulate and compare different forecast scenarios through 2023. The most likely scenario results in a forecasted need of approximately 5.3--6.3 million skilled laborers needed in the construction industry by 2023.;Recommendations are given as to how the availability and quality of construction industry labor data can be improved. One key recommendation is that more extensive data collection can be undertaken to produce accurate and consistent data. Doing so will provide support for more accurate forecasts for planning, recruitment, and retention efforts. Also, the industry should use and strive to improve the newly developed metric for construction industry labor productivity, since this allows construction professionals to be able to analyze industry level productivity by means of a commonly used industry reference manual.;Overall, the research findings present a reliable forecast model that produces short to medium term forecasts of skilled labor demand with reasonable accuracy. Construction industry participants and stakeholders, including practitioners, owners, researchers, training providers, government agencies, and employment policy makers, can use the resulting data, model, and forecast scenarios to be proactive in their planning and policy making as it relates to ensuring an adequate skilled construction labor force in the future.