The Olympic optimism bias has left the taxpayer out of pocket


There is a myth that the London Olympics were delivered within budget. This claim was used recently by David Cameron, the British prime minister, to win support for the proposed high-speed railline to Birmingham.

The principal relevant facts are these. The first detailed specification of what was needed for London to host the games was drawn up in 2002 by Arup. The report by the engineering and planning consultancy put the cost at £1.8bn, much of it to be privately financed. An extended assessment was then commissioned by the Department of Culture, Media and Sport from PwC. The financial consultancy’s 2003 report estimated the total cost at £3.1bn, requiring a public subsidy of £1.3bn. The balance would be recovered from the private sector and from asset sales after the games. According to PwC’s risk assessment, the probability that the taxpayer would need to provide as much as £2bn was less than 5 per cent.

The detailed bid submitted to the International Olympic Committee in 2005 set out a total cost of £4.2bn. In accepting London’s proposal, the IOC commented that the operating cost budget seemed realistic; and that, since the British government had committed to underwrite expenditure on infrastructure and facilities, it was unnecessary for the IOC to review the cost assumptions further.

It was nevertheless necessary for the DCMS to review the cost assumptions further. The budget had by 2007 increased to about £6.5bn. At Treasury insistence, a contingency allowance, mostly unspecific, of £2.8bn – more than the total original projected cost – was added. The costs of land acquisition and of the Olympic Village were mostly excluded in the belief that they would be recovered from property sales after the event.

The basis of the claim that the games came in “within budget” seems to be that a small part – currently £300m-£400m – of that £2.8bn contingency remains unspent. That was achieved, however, by excluding a number of additional unbudgeted expenditures from the calculation, as the National Audit Office has highlighted. There is likely to be little, if any, net recovery of the further costs of land and housing, which were due to be recouped from property sales.

The costs were grossly and persistently underestimated, and the financial contributions anticipated from private sources overestimated by very large amounts. Every year, to the present day, the expected cost rose and the likely revenues diminished. The cost of the games to public funds has proved to be about 10 times the original estimate.

There is a halo effect; the sporting success of the Olympics fosters the mistaken belief that they were an economic success. Papers recently produced by government on the “economic benefits” must be an embarrassment to the many good people of the Government Economic Service, conflating incommensurable monetary amounts and confusing costs with benefits. At the same time, a curious puritanism requires politicians to pretend activities intended to make us feel good about ourselves are justified by their contribution to “the economy”. The Olympiad was a good party, which cost the British population about £200 per head.

False accounts of the past prevent us learning lessons for the future, of which there are many. The Olympics remind us that enthusiasts typically understate costs and overstate benefits. Consultants win work by pleasing clients, and they rarely please clients by pouring cold water on their pet schemes. We should waste no more public money on risk simulations such as those in the PwC report; the outcome in this case was one of these supposedly statistically impossible events that seem nevertheless to occur on a daily basis. Establishing an allowance for “optimism bias” is realistic, but offers little incentive to make careful projections in the first place. Money put in a budget for general contingencies is not money you are likely to see again.

Overruns on the 2012 Olympic Games were no worse than those in other host cities from Sydney to Rio. Misleading estimates of costs and revenues are endemic to large, idiosyncratic public sector projects. However, the work of Professor Bent Flyvbjerg associates at Oxford university identifies sporting events and rail schemes as among the worst offenders. Surely there is a lesson there for Mr Cameron.

Print Friendly, PDF & Email