The winner's curse and construction tenders

Finally, some sanity in terms of writing about the problems the construction industry is facing. John Walton wrote in the New Zealand Herald earlier this week:

Right now, of course, the construction industry is indeed booming - not just in housing, but also in commercial and infrastructure development. But companies are continuing to fail. Why?

The answer lies in the construction process and a misunderstanding of the roles of the participants. At its simplest, the owner provides the site, resource consents, designs and pays for the work. The contractor organises the work to the design and to the required legal standards, for the agreed price within the allocated time. Price and time are adjusted for unforeseen events and for changes instructed by the owner. To this extent, construction contracts legislate for uncertainty.

That uncertainty is exacerbated by an incomplete understanding of other project risks (ground conditions and supply chain issues like subcontractor and supplier pricing and availability) and unrealistic expectations on the part of owners, particularly that they can fill in the gaps in the design and instruct changes at their whim without cost consequences. The tender process encourages this opportunistic behaviour by forcing contractors to compete on incomplete, or simply unrealistic or unfair contract terms.

Contractors try to introduce some balance by excluding risks from their bids. They must then rely on the claims process to protect their margins.

This can turn the pricing process into something of a lottery. Typically, the cheapest price wins, which all too often is submitted by the contractor with the greatest appetite for risk, coupled with the most optimistic expectations for making claims under the contract.

Following contract award, managing design development, construction and capricious owner changes to the design becomes a considerable headache for contractors who need to be able to meet construction costs, pay subcontractors and protect their already slim margins.

Every time the issue of financially troubled construction companies comes up (and it seems to be coming up a lot lately), it makes me think of the winner's curse.

Consider a group of construction firms, tendering for a construction contract. Rational construction firms who have complete information about the project and associated risks (and with the same tolerance for risk) would all have the same expectations about the costs of completing the contract, so all would bid the same. However, not all construction firms have complete information (as Walton noted above) and not all firms have the same tolerance for risk. The construction firms may make random errors in determining their costs (or margins) and risks associated with the contract, so all of the construction firms will expect different completion costs (and margins and risks) associated with the contract. For firms with similar tolerance for risk, differences in expected completion costs (and margins) arise randomly - some will overestimate the completion costs (or underestimate the risk) of the contract, and some will underestimate the completion costs (or overestimate the risk or margins). Those who expect low completion costs will bid low for the contract, and those who expect high completion costs will bid high for the contract.

The real problem arises when the contract goes to whichever construction firm bids the lowest for the contract. This will be the firm that has most underestimated the completion costs, because they will be the firm that bid the lowest. The chances are high that, if there are enough construction firms entering bids, the eventual winner will have underestimated the completion costs compared with the true costs, and hence will not receive enough to cover their true costs. This is what we refer to as the winner's curse.

The problem is that consumers of construction firms' services are too focused on looking for the lowest bid. This virtually guarantees the problems we are facing in the construction industry. Walton's piece concludes:

The industry has a choice. Either it accepts that designs and prices will change and pay contractors accordingly, or take the time to remove contract uncertainties before fixing the price and instructing work to commence. Experience here and overseas would suggest that a combination of the two works best.

Walton is not very clear on his proposed solution. So, let me offer two options.

First, when a construction contract is put up for bids, the decision could be made independent of price. That removes the incentive to bid too low. Construction firms can then be realistic about the costs and risks they face, when preparing their bids, without worrying about a high bid ruling them out. Of course, it probably creates an incentive to bid too high, precisely because a high price won't rule the firm out of the process.

Second, adopt a variant of a second-price auction. Give the contract to the lowest bidder, but pay them the amount that was asked by the second-lowest bidder. Or, if the contract is not given to the lowest bidder, still pay an amount for the contract equal to the next highest bidder's offer. This ensures that the client isn't paying well over a 'fair' price for the contract (which would likely be the case for the first option), while providing some additional space for the successful firm, which has likely underestimated the completion costs. If second-price isn't enough, a third-price auction would further limit the chance of construction firms being underpaid because of their inability to accurately estimate costs.

Something clearly needs to be done. We don't want construction firms falling over mid-contract, and in a small market like New Zealand we can't afford to have the market dominated by only a couple of players. We need to ensure that sustainable contract prices are being paid, and the current system is clearly failing.