Measuring and Modelling Overruns in Highway Project Costs in Nigeria

Abstract

Overruns in highway project costs are pervasive globally. Recent attempts at predicting highway cost overruns predominantly rely on the use of after-the-fact machine learning predictive models that mainly address developed country contexts. The present study estimated the extent of cost overrun in Nigerian highway projects, and developed a regression model for predicting cost overruns in such projects based on the winning tender. Secondary data for the study was obtained from the website of the Nigerian Federal Ministry of Works and Housing, and covered the period 2002 to 2015. The overall average cost overrun for Nigerian highway projects was determined to be 17.37%. Additionally, a logarithmic model for estimating cost overrun at the pre-award stage of highway projects was obtained as 𝐿𝑜𝑔10(𝐶𝑜𝑠𝑡𝑜𝑣𝑒𝑟𝑟𝑢𝑛) = 0.844 + 0.914𝐿𝑜𝑔10(𝑖𝑛𝑖𝑡𝑖𝑎𝑙𝑐𝑜𝑠𝑡). In the context of this study, it is better to estimate overruns in highway projects after the successful bidder has been identified, and to add the estimated cost overrun as a contingency sum to the contractor’s bid at that stage. Measures for mitigating cost overrun in highway projects in Nigeria should be intensified