Rethinking Construction Cost Overruns: Cognition, Learning and Estimation

Dominic Ahiaga-Dagbui, Simon Smith

Research output: Contribution to journalArticlepeer-review

Abstract

Cost performance on a construction project remains one of the main measures of the success of a construction project (Atkinson, 1999; Chan and Chan, 2004). Reliable estimates are important for several reasons- for organisational budgeting purposes, for loan application if project has to be funded through credit facilities, to estimate their likely cost of financing loans (interest payments), for estimate commercial feasibility or viability of the project, etc. The present economic meltdown also imposes a parsimonious approach to spending on most organisations and governments. However, estimating the final cost of construction projects can be extremely difficult due to the complex web of cost influencing factors that need to be considered- type of project, material costs, likely design and scope changes, ground conditions, duration, size of project, type of client, tendering method- the list is endless (Ahiaga-Dagbui and Smith, 2012). Trying to work out the cost influence of most of these variables at the inception stage of a project where cost targets are set can be an exhaustive task, if not at all futile. Ignoring most of them altogether creates a perfect recipe for future cost overruns, disputes, law suits and even project termination in some cases. Even more, there is a high level of uncertainty around most of these factors at the initial stages of the project as noted by Jennings (2012).
Original languageEnglish
Pages (from-to)38-54
JournalJournal of Financial Management of Property and Construction
Volume19
Issue number1
DOIs
Publication statusPublished - Apr 2014

Keywords / Materials (for Non-textual outputs)

  • Data mining, Prospect theory, Cost overruns, Dunning-Kruger effects, Optimism bias, Referenced class forecasting

Fingerprint

Dive into the research topics of 'Rethinking Construction Cost Overruns: Cognition, Learning and Estimation'. Together they form a unique fingerprint.

Cite this