The Treasury

Global Navigation

Personal tools


Guide to Social Cost Benefit Analysis

Appendix 3: Example CBAs

297.This appendix describes two CBAs to illustrate some of the difficulties encountered in estimating costs and benefits and some of the methods used to overcome them.

Bid to host the 2018 Commonwealth Games in Auckland


$m NPV

Quantified Benefits Notes Option 1
Minimum price
New Zealand attendees (a) $71m
Induced tourism (b) $22m
“Halo” effects[26] (c) $105m
Infrastructure (d) $930m
Television revenue (e) $60m
Sponsorship, etc (f) $45m
Total   $1,232m
Quantified Costs [27]    
Bid budget (g) $21m
Infrastructure (g) $930m
Operating costs (g) $835m
Other capital (g) $160m
Total   $1,945m
Net present value   -$713m
Benefit cost ratio   0.63
Unquantified Benefits    
National pride (h)  
International profile (h)  
Business opportunities (h)  
Social, cultural and community (incl volunteering) (h)  
Sports development (h)  
Event culture (h)  
Physical regeneration (h)  
Unquantified Costs    
Congestion externalities (h)  
Risk (h)  

298. The CBA was undertaken by LECG (now trading as Sapere Research Group) with input from Covec Ltd and NZIER in 2009/10. The CBA applied a discount rate of 8% pre-tax, real.

Analytical issues

299.Counter-factual: a lot of care was taken to separate those events and effects that were attributable to staging the event, from those that would occur anyway, or that were not necessary to stage the event.

300.A 4% discount rate was tested, but was found to have only a negligible effect on the benefit-cost ratio.

301.A national rather than a regional perspective was taken, and foreign consumers were not given standing.


302.A computational general equilibrium (CGE) analysis was carried out. This captured flow-on benefits and costs of the spending on infrastructure, running costs and increased tourism related to the games. It incorporated displacement effects and resource constraints, and the phasing of expenditure and debt repayment. It found that the games generate strong positive benefits for Auckland over the course of the games development, as well as net gain over the long term. However at the national level, the games generate positive benefits in the short term as expenditure on construction and by tourists boost the economy, but result in a welfare loss in the long term, as funding is repaid. This reinforces the conclusion of the cost benefit analysis, although it should be noted that the CGE does not capture some of the quantified benefits in the cost benefit analysis (such as the ‘halo’ effect), and none of the unquantified benefits and costs.


303.Benefits estimation

  1. The report referred to this as “in-person spectacle” (mainly ticket sales): to determine the consumer surplus, a demand curve was assumed based on an estimate made in a CBA of the 2010 Vancouver Winter Olympics.
  2. Tourism/visitation impacts: Specific modelling work was undertaken using case study material from Melbourne 2006 and ‘business as usual’ modelling of Auckland-specific visitation patterns. An economic impact analysis was carried out which was then translated into a cost-benefit framework. Account was taken of displacement effects (‘crowding out’ of usual tourists/visitors) and the effect of expected price changes.
  3. 'Halo' effect: This is known in the literature as an existence (or non-use) value. In general, stated preference valuations are used to determine existence values, again using the concept of willingness to pay. However, this was not done. Instead, results of contingent valuation surveys carried out in London, Manchester and Glasgow to measure willingness to pay for the London Olympics were used as a starting point, and adjusted for the New Zealand context. This is referred to as the ‘benefit transfer' method. Benefit transfer methods are highly problematic, as discussed in this guide, so the study should have included some discussion of the validity of those surveys. Moreover, it is not clear how some of the adjustments were arrived at. For example, it was assumed that Auckland citizens would pay at most around half of what London citizens said they would pay to host the Olympics, but no explanation is given for why this should be so.
  4. Infrastructure benefits: The report noted that
    • many new road infrastructure projects in New Zealand have expected benefit-cost ratios in excess of 1
    • most infrastructure required for the games is not roads, and
    • the only infrastructure that would not have been built anyway is infrastructure that in the period after the games, or in the absence of the games, would likely have benefit-cost ratios of less than 1.

    The report therefore assumed that on average, benefit-cost ratios of infrastructure are equal to 1, with a high scenario of 3.15 and a low scenario of 0.5.

  1. Television spectacle: most of the benefits of being able to watch the games on television were assumed to go to foreigners, which have no standing in this analysis. Also, the benefit to New Zealanders of watching the games on television was not counted, on the grounds that they would get that benefit regardless of where the games were held. That benefit could therefore not be attributed to holding the games in New Zealand. The value of the televised spectacle is equal to the revenue derived by the host. However, the report provides no information on how the figure of $60 million was worked out: it appears that it represents an estimate made by the Central/local Government Bid Feasibility Group.
  2. Sponsorship income: the $45 million is based on an estimate made by the Bid Feasibility Group.

304.Costs estimation

  1. Infrastructure and other costs were obtained from the Central/local Government Bid Feasibility Group’s estimates. It appears that no assessment was made of whether there was an optimism bias.

305.Unquantified costs and benefits

  1. Unquantified benefits: the report discussed these at length. For many of these, it concluded that the benefits were transient or ephemeral, or constituted double counting of other concepts (for example, national pride at least partly overlaps with the ‘halo’ effect).


  • [26]Sense of enjoyment from hosting the CWG
  • [27]These include a 20% deadweight cost for taxpayer-funded expenses, as recommended in the Treasury’s CBA primer.
Page top