Ideas and Impact
The School

Olympic proportions: cost and cost overrun at the Olympics 1960-2012

Bent Flyvbjerg and Allison Stewart

Saïd Business School Working Papers, Oxford: University of Oxford, 23 pp.

Do different types of megaprojects have different cost overruns? This apparently simple question is at the heart of research at the University of Oxford aimed at understanding the characteristics of megaprojects, particularly in terms of how they are established, run and concluded. In this study, we set out to investigate cost overruns in the Olympic Games.

To do so, we examined the costs of the Games over half a century, including both summer and winter Olympics. We looked at the evolution of final reported costs and compared these to the costs established in the Games bids, submitted to the International Olympic Committee (IOC) up to seven years before the Games occurred. In so doing we established the largest dataset of its kind, and documented for the first time in a consistent fashion the costs and cost overruns for the Olympic Games, from 1960 to 2012.

We discovered that the Games stand out in two distinct ways compared to other megaprojects:

(1) The Games overrun with 100 per cent consistency. No other type of megaproject is this consistent regarding cost overrun. Other project types are typically on budget from time to time, but not the Olympics.
(2) With an average cost overrun in real terms of 179 per cent – and 324 per cent in nominal terms – overruns in the Games have historically been significantly larger than for other types of megaprojects, including infrastructure, construction, ICT, and dams. The data thus show that for a city and nation to decide to host the Olympic Games is to take on one of the most financially risky type of megaproject that exists, something that many cities and nations have learned to their peril.

For the London 2012 Games, we find that:

(1) With sports-related real costs currently estimated at USD14.8 billion, London is on track to become the most costly Olympics ever.
(2) With a projected cost overrun of 101 per cent in real terms, overrun for London is below the historical average for the Games, but not significantly so.
(3) The London cost overrun is, however, significantly higher than overruns for recent Games since 1999. London therefore is reversing a positive trend of falling cost overruns for the Games.

Bent Flyvbjerg
Allison Stewart