The failure to improve in costs estimate over the decades is attributed by Flyvbjerg et al to deceit and bias in initial costings to get projects started.
This reasoning is problematic. Much of both the organisation of markets, business practices and the internal organisation, size and shape of firms emerged as the result of efforts to curb moral hazard and principal-agent conflicts, to reduce transaction costs and uncertainty, and to spur innovation and individual initiative. Part of the reward to entrepreneurs for owning a firm is from their successful policing of the efforts of employees and sub-contractors they have hired.
The correcting of errors in cost estimates and the subsequent building of mega-projects are entrepreneurial opportunities for pure profit. The issue is how to organise markets, firms and procurement contracts to correct these errors.
For example, over the last few decades, private sector mega-projects promoters developed their own special form of financing known as project financing (or non-recourse debt) to improve incentive structures and risk allocation.
Major engineering projects can generate a good reputation for the consultants involved. They can also damage the image of the whole profession if things go wrong, and things most clearly go wrong when the final cost of the project is much higher than the initial estimate.
In theory this should be a random and infrequent occurence, particularly if consultants are as good as they claim to be. However a study by Flyvberg in 2005 showed that “Overrun is constant for the 70-year period covered by the study, estimates have not improved over time” (ref. 187 page 3).
Examples
Cost overruns can be large. For example, for large rail projects, cost overruns can range up to 44% and with forecast rail passengers over-estimated by over 105% (see 187 page 3). These figures must assume that everyone involved can agree on the definitions of costs and benefits, and on how to evaluate…
View original post 592 more words
Utopia, you are standing in it! 
Recent Comments