Abstract
The debates on whether to bid for organizing a mega sports event like the World Cup Soccer or the Olympic Games ignore either the bidding costs or the probability not to win the bid or both. In this short article, I discuss why the bidding costs and probabilities should be taken into account and I show for the Netherlands, which is currently discussing bidding for the Olympic Games of 2028, that the Net Present Value of organizing should be €557 million to €2.8 billion positive to compensate for the bidding costs and the probability of an unsuccessful bid.
Introduction
In the Netherlands, there are serious plans and discussions about bidding for the Olympic Summer Games of 2028. The organization “Olympic Flame” is lobbying for this, supported by among others the Dutch Olympic Committee. The Dutch government seriously considers supporting such a bid. A decision on whether to support the bid is likely to be taken in 2016, because then there is still ample preparation time for a bid to be handed in 2019. The government is now alreAdy studying how to organize the Olympic Games, if they are awarded to a Dutch city, because planning and realizing large infrastructural procedures requires substantial time in the Netherlands. Waiting with preparations until 2016, 2019, or 2021 when the International Olympic Committee (IOC) decides on which city organizes, might leave problematic little time to organize the Games.
As part of the preparation, the Dutch government (the Ministry of Health, Welfare and Sport) commissioned a social cost–benefit analysis (SCBA). This SCBA is rather explorative because the plans are still broad sketches and lack detail yet. The SCBA includes seven ways of organizing the Games in the Netherlands. This SCBA, Blok et al. (2011), concludes that the net present value in 2020, depending on the alternative, varies between −€1.1 and −€1.8 billion in the likely scenario. Blok et al. indicate that better alternatives are possible, which could have a positive outcome.
In De Nooij (2012), 1 I argue that Blok et al. are too positive. Blok et al. (2011) calculate the bottom-up cost, so what investments are needed and how much will those investments cost. Previous Olympics and other major sports projects have shown that this type of cost estimates often result in a substantial underestimation of the actual costs, and an overestimation of the benefits. See for example, Crompton (1995, 2004), Hudson (2001), Matheson (2002, 2006), and Preuss (2004, p. 290). Based on the experience of Athens (2004) and London (2012), the investment cost increase (double or more than that), as do the costs of security compared to Blok et al. De Nooij (2012) therefore shows that the probable negative balance is −€4.0 billion. The bandwidth for this estimate is wide: from −€10.3 to €2.0 billion.
And yet, both Blok et al. (2011) and De Nooij (2012) are too positive and not suitable for the decisions politicians now face. Both studies analyze the net present value for the Netherlands of organizing the Games once the Netherlands are assigned the Games in 2021. For now, the choice is rather “the Netherlands is continuing with the preparations for the bid and the government keeps spending money and resources until there is a decision in 2016 or 2021.” The choice is not “should the Netherlands organize.” This preparation costs before the decision is taken are included in a limited way in both studies. More importantly, both studies do not take into account the probability of success, or actually the probability of no success. In this article, I show how this can be corrected for. This correction is not only relevant for the Netherlands but for all countries considering to prepare a bid. I am not aware of any study already showing such a correction.
Calculation of the Expected Welfare Cost of Bidding for the 2028 Olympics for the Netherlands
To calculate the positive present value of organizing the Olympic Games (x) that compensates both the preparation costs and the probability of success the following equation should be solved:
where b is the bidding cost and p is the probability to win the bid. Solving for x gives
To calculate this compensating net present value of organizing the Olympic Games (x) both the preparation costs and probability should be known. This section discusses both, followed by the calculation of the necessary net present value of organizing to make bidding (or considering to bid) economically attractive.
Preparation Costs
There are no hard figures on how much it costs to decide to bid and how much preparing the actual bid costs. There are a few numbers around, but it is unclear what they include and what they do not include (see also the concluding section). For the Netherlands two estimates are available, which in itself is already a nice addition to the literature.
The first estimate is from Blok et al. (2011, p. 53), who, in their SCBA include the preparation and bidding costs for the government. In their base case 2 they assume that governments (state and several municipalities) annually spend 10 full time equivalent (FTE) in the preparation, with a government FTE costing including overhead about € 100,000. They assume that out of pocket costs (mostly consultants) double this. This makes that the annual costs are € 2 million from 2012 to 2016 when the government decides whether to support a bid, and the total costs in this phase are € 10 million. They estimate the costs to actually prepare a bid at € 42 million. They estimate this on two estimates for the Chicago 2016 Olympic bid. They report that one source says that the Chicago bid had cost € 37 million cost, while another source reported a cost of € 79 million (they do not give the references).
The second estimate is from RTLnews, which is the largest commercial Dutch broadcaster. Following the publication of Blok et al. RTLnews investigated how much the government already had spend. For this they asked municipalities, provinces, and the central government how much they had already spend and what their plans were. Table 1 lists all the expenditures they found. They reported that already € 190 million was spend. € 81 million they classified as direct costs for the Games in 2028, such costs feasibilities studies and the costs of lobby group ‘Olympic Flame’ (Olympisch Vuur in Dutch). The remaining € 109 million are indirect expenditures and go to new and improved sports facilities at Olympic level. These new facilities are not necessarily the facilities to be used during the 2028 Olympic Games if these are in the Netherlands.
Expenditures for the Dutch 2028 bid so far.
Source. RTLnews (2012)
However, their conclusion that already € 190 million is spend is tricky for three reasons. 3 Firstly, not all amounts are actually for the decision to bid for the Olympics. Some of these amounts are for the Olympic Plan which among other things aims to increase sports participation in the Netherlands and another part for accommodation which even without Olympic Games have a value. Secondly, it is not an annual amount. Roughly one third of the expenditure (€ 68 million) is spend by the Ministry of Health, care and sports. This amount consists of € 11 million in feasibility studies and € 2 million in subsidies, both annually up to 2016. Thirdly, it is not complete because the time of civil servants is not included nor valued. (The question is whether the opportunity cost of the officials is included in the cost estimate for Chicago).
The Probability
To calculate how much a country should gain if it’s bid is chosen, depends on the probability its bid is (not) chosen. This requires predicting the future, and especially in the case of the Netherlands considering to bid for 2028 the relatively far future. Asking those involved about their (subjective) estimate of the probability of success is unlikely to give reliable estimates. The approach I follow here is to calculate the average probability for a city, considering to bid, to actually issue a bid; the probability a bid is chosen on IOCs shortlist; and the probability a bid on the short list is chosen as the winning bid. These different probabilities matter for the welfare calculation because each success increases the costs (because preparation goes on longer). This method ignores what are the other cities that bid and the quality of the respective bids, it ignores who organizes the Games in 2024. So for a specific bid the estimate might be wrong, but on average it is correct.
Table 2 shows for four Olympic Games how many cities considered to cast a bid but did not, how many cities have cast a bid and how many bids are included in IOC’s final consideration. Thus, for 2012 IOC received 9 bids. Of these, five came on the short list for further consideration, and ultimately London was chosen. There were about 12 cities considering to cast a bid but that did not cast a bid. So, in total, 21 cities were considering to cast a bid, and only one city was awarded the Games. For 2016, the IOC received 7 bids and four shortlisted, and ultimately chose Rio de Janeiro. There were over 17 cities mentioned as potential organizers that did not case a bid. For 2020 5 cities still compete for the honor to organize the Olympics. One bid is withdrawn (Rome). 10 cities had said to bid but had not formally announced, and ultimately did not. 10 cities had formally announced to bid but did not cast a bid. For 2024 now 21 cities are mentioned, which is a provisional number because the application deadline is about three years from writing this article.
Number of (potential) bidders for organizing the Olympic Summer Games.
Source. Wikipedia (2012a)
Calculation for the Netherlands
Using the cost and probability figures above a calculation of how much the net present value of organizing the Olympic Games should be to compensate for the costs made if lost or not cast a bid after seriously discussing this. Because of the uncertainty about the cost estimate two calculations are presented. In terms of probability of taking the next step or having success I am assuming the average for 2012, 20126 and 2020: 16 cities have expressed interest to cast a bid but did not actually bid, 7 cities bid and from these 5 cities are selected by IOC for the final selection round and one city is awarded the Games. The probability of going from ‘considering to cast a bid’ to ‘casting a bid’ is 7/23, the probability to reach the short list is than 5/7. The opportunity to be successively be selected by the IOC is 1/5. The probability of success from the moment of considering to cast a bid is 1/23. 4
If the cost estimate of Blok et al. is correct, then during the first five years the cost are € 2 million annually, while after a positive decision to cast a bid the annual costs are € 8.4 million. Adjusted for the probability a next stage is reached, the expected net present cost in 2012 of launching a bid are € 15.8 million. To compensate, there should be net benefits of organizing the Olympic Games. However since these benefits are only there in case of success (organizing), while the costs will always materialize, the net benefits of organizing should be much larger. Given the probability to win is 1 in 23, the net present value of organizing the Games should be € 363 million in 2012 Euros or € 557 million in 2020 Euros. The calculation is illustrated in table 3. Blok et al. and De Nooij discount to 2020, what is the moment the Games are (almost) awarded. For comparability reasons the net present value is calculated for 2020.
Calculation of the minimum required net present value of organizing the Games excluding bidding costs to make bidding economically attractive.
Note. In the calculations the standard discount rate for Dutch government projects of 5.5% is used.
If the estimated annual costs are higher, for example at € 15 million, which based on the RTLnews data seems realistic, then by a similar calculation the present value of organizing the Games in 2028 should be € 1.8 billion in Euros of 2012 or € 2.8 billion in Euros of 2020 to compensate for the bidding costs and the probability of not winning the bid.
Conclusion
The economics of organizing a mega sport event like the World Cup Soccer or the Olympic Games is not favorable, but the economics of the bidding to organize such a mega event is even worse. In the previous section I calculated that for the Netherlands the NPV of organizing the Olympic Games should be between € 557 million and € 2.8 billion (2020 price levels) positive to compensate for the costs if the bidding process does not result in organizing the Games.
Calculations can be broadened to other events and countries. For example, a Dutch newspaper reported that an Australian senator demanded € 34 million back from FIFA after not winning the World Cup 2018 or 2022 after rumors the election was not fair. (AD 30/05/11). 5 Given that when bidding for both tournaments started 11 bids were handed in (one which was withdrawn and one that was rejected) for both tournaments, the probability of winning is about 18% (Wikipedia, 2012b). The bidding probability compensating value of organizing the World Cup (x) following equation (2) is thus € 187 million (34/(2/11)). The Dutch bidding organization (the HollandBelgium Bid, 2011) reported bidding costs of € 10.4 million. However this excludes the costs of the government (state and municipalities), the stadiums and support in kind. Unclear is how high these total cost are and whether they are comparable to the Australian costs.
The city of Chicago, one of the seven cities that issued a bid for the 2016 Olympic Games, is said to have spent $ 100 million on its bid (Zimbalist 2010, p. 8). The present value of actually organizing the Olympic Games that compensate for bidding costs and the bidding probability (x) according to equation (2) should have been at least $700 million (100 / (1/7) = 700) if the Games would have been awarded to Chicago.
The value of winning a mega event to compensate for the bidding costs will vary over time and between countries. The costs of preparing a bid will vary between countries and times. As does the probability of being awarded the event, which varies with the numbers of bidders, with on the one extreme the Summer Olympics of 1984 for which Los Angeles was the only bidding city (Preuss, 2004, p. 7).
For the Netherlands, continuing with discussing whether to bid and preparing a bid seems impossible to be underpinned using economic arguments. Continuing to investigate how a bid should be designed such that it is attractive for both the Netherlands and the International Olympic Committee should therefore be based on the desire to host and to celebrate the Games.
And in a broader perspective, ignoring the probability of no success and ignoring the bidding costs in most of the cost–benefit analysis of organizing a mega sports event serious overestimates its economic attractiveness. This is a further argument why many economic appraisals are too positive.
Footnotes
Acknowledgments
This article benefited from discussions and comments on earlier drafts from Gelijn Werner, Marcel van den Berg and an anonymous reviewer. The author is solely responsible for any remaining errors.