Cost forecasting using RCF: a case study for planning public building projects costs in Turkey

Savas Bayram* (Corresponding Author), Saad Al-Jibouri

*Corresponding author for this work

    Research output: Contribution to journalArticleAcademicpeer-review

    10 Citations (Scopus)


    Many construction projects are perceived to perform badly since the majority of them suffer cost overruns. This paper analyses cost estimates produced by two most commonly used estimation methods for building projects in Turkey, namely unit area cost (UAC) and unit price analysis (UPA). It also investigates whether reference class forecasting (RCF) method can be used to provide the client with more certainty in the final project cost forecasts compared to those two methods. Data from 369 public building projects in Turkey were collected and cost overruns were analysed. Part of the data representing 14% of the projects were randomly selected and used as control group for testing the reliability of the forecasts produced using RCF. The results have shown the estimates of the final project cost produced using the UAC and UPA methods in Turkey are either too pessimistic or optimistic. The results have also shown that it is possible to use the reference class as a basis to produce a reasonably accurate prediction of the actual cost of a project in a similar class for various levels of acceptable risks. This is achieved by adding uplifts/reductions to the estimates produced by the two traditional methods.

    Original languageEnglish
    Pages (from-to)405-417
    Number of pages13
    JournalInternational journal of construction management
    Issue number5
    Publication statusPublished - May 2018


    • optimism bias
    • reference class forecasting
    • unit area cost
    • unit price analysis
    • Cost forecasting
    • n/a OA procedure
    • NLA


    Dive into the research topics of 'Cost forecasting using RCF: a case study for planning public building projects costs in Turkey'. Together they form a unique fingerprint.

    Cite this