Forecasting Ratio of Low Bid to Owner’s Estimate for Highway Construction

Abstract

Several state departments of transportation (state DOTs) have encountered significant challenges to accurately estimate costs of their highway projects. It is not uncommon that states’ DOT estimates (owner’s estimates) are significantly different from contractors’ submitted bids. This is a critical problem for state highway agencies that strive to develop more accurate cost estimates, deliver projects within the budget, and optimize constrained funds for their highway programs. This inaccuracy problem is a temporal issue since the engineer’s estimate is developed ahead of time before the project is advertised and bids are received. The question that transportation agencies are interested in finding an answer to is: are there any significant risk factors in the construction market indicating to the increased likelihood of the deviation between owner’s estimate and the submitted low bid? In this research, a temporal perspective is selected to answer this question through identifying risk factors affecting the accuracy of the owner’s estimate. The ratio of low bid to owner’s estimate is examined using time-series analysis. The objectives of this research are to (1) examine several variables representing local highway construction market, overall construction market, macroeconomic conditions, and energy market, to identify the leading indicators of ratio of low bid to owner’s estimate; and (2) use the identified leading indicators to develop an appropriate time-series model to forecast the ratio of low bid to owner’s estimate. Four variables are identified as the major leading indicators: (1) number of projects awarded in the same month at the state level; (2) average number of bidders last month; (3) producer price index for steel mill products (PPISM); and (4) construction cost index (CCI). Several seasonal autoregressive integrated moving average with explanatory variable (ARIMAX) models are developed that are capable of forecasting the ratio of low bid to owner’s estimate with a high accuracy. This research contributes to the state of knowledge of analyzing the difference between owner’s estimate and low bid through: (1) identification of leading indicators of ratio of low bid to owner’s estimate that convey the extent of risk and uncertainty associated with construction projects at the cost estimation phase; and (2) development of appropriate multivariate time-series models (i.e., ARIMAX models) to predict the ratio of low bid to owner’s estimate. It is anticipated that the results will help cost estimating professionals in transportation agencies better understand the variability between their estimates and submitted bids by highway contractors, and thus, prepare more accurate cost estimates and develop appropriate risk management strategies for enhanced decision-making.