Analysis of the Determinants of Cost Overruns in Road Construction Projects in Tanzania: A Probit Regression Model Approach

Abstract

Road construction projects in Tanzania continue to experience persistent cost overruns, undermining infrastructure development and broader economic growth. This study employs probit regression analysis to identify the key factors influencing cost escalation in road construction projects. The main objective was to determine the determinants of cost overruns by examining seven categories of influences: client-related, consultant-related, contractor-related, project-related, government-related, managerial, and economic factors. The study focused on major road sector stakeholders, including contractors, consultants, and clients affiliated with TANROADS and TARURA. A cross-sectional research design was adopted, with primary data collected from 200 respondents using stratified and simple random sampling techniques. The final sample comprised 100 contractors, 50 clients, and 50 consultants. A total of 179 valid responses were obtained, representing a 89.5% response rate. Respondents were senior professionals with substantial industry experience, and the sampling frame was drawn from the Contractors Registration Board (CRB) database. Probit regression results indicate strong overall model performance, with a likelihood ratio chi-square statistic of 66.16 (p < 0.0001) and a pseudo-R² of 0.3179. Consultant-related factors (dy/dx = 0.8091, p < 0.001) and managerial inefficiencies (dy/dx = 0.7834, p < 0.001) were found to significantly increase the probability of cost overruns. Client involvement also exhibited a notable effect (dy/dx = -0.3705, p = 0.079). In contrast, government-related factors demonstrated a significant negative association with cost overruns (dy/dx = -0.5259, p < 0.001), suggesting the effectiveness of recent reform initiatives. Project-related, contractor-related, and economic factors were not statistically significant. The study concludes that improving consultant performance, strengthening managerial capacity, and enhancing governance mechanisms are critical for effective cost control in Tanzania’s road sector. This study contributes to the existing body of knowledge by providing empirical evidence on the relative importance of consultant-related, managerial, client-related, and government-related factors in explaining cost overruns in Tanzanian road construction projects. Unlike many previous studies that relied primarily on descriptive analyses, the study employs a probit regression approach to quantify the probability effects of different determinants, thereby offering context-specific insights for infrastructure policy and project management in Tanzania.