The purpose of this paper is to make some observations from the perspective of a project lawyer on the Canadian law of tender as it has developed over the past 30 years since the Supreme Court of Canada’s seminal decision in R. v. Ron Engineering & Construction (Eastern) Ltd. (“Ron Engineering”).1
It has been suggested that “[t]he law of bidding and tendering is a juxtaposition of the principles of fairness and transparency in the tender process and the freedom of parties to contract upon terms that they negotiate or agree upon.”2 This view imputes to the tender process a fairness requirement that is generally not found in the law of contract in Canada. This fairness requirement is seemingly defined by reference to an idealized vision of tendering that apparently exists beyond the four corners of a set of tender documents. Some proponents of this view also advocate for an ancillary requirement of “transparency” in the common law of tender, supposedly so that tenderers can have confidence in the fairness of the evaluation of tenders and the selection of the successful tenderer.
The thesis of this paper is that such a juxtaposition is unnecessary, and, in fact, is causing unanticipated harmful side effects for tender processes. Tendering is simply a vehicle by which parties enter a private contract. Based on the authors’ experience there is no compelling reason why a tender process in Canada should attract a separate duty of fairness or a requirement of transparency beyond that which the general law of contract provides. In fact, in our observation, to the extent Courts have implied into Contract A additional requirements of fairness and transparency in tendering beyond what was reasonably anticipated by the participating parties at the time of the tender closing, little has been achieved to “protect the integrity of the tender process”. Rather, this progression away from a strictly contractual‐based view of the law of tender is causing uncertainty and additional costs to both owners and tenderers, and ironically is providing opportunities for mischief, thereby promoting the very harm that the Supreme Court of Canada was trying to prevent in Ron Engineering.
In some decisions there are statements that have a flavour of “consumer protectionism”, apparently based on a view that in a tender process there is an inherent imbalance of power between an owner and tenderers which the Court should adjust by the imposition of extracontractual duties. Based on our experience there is no such imbalance. That view ignores commercial realities, including the usual steps an owner follows in deciding to invite the market to enter a contract and in drafting tender documents. The owner does not take such steps in a vacuum. However, even if there were such an imbalance, in this commercial field, unlike in the consumer protection area and other areas where there is a concern for individual rights related to concentrated economic power, we submit it should be for the legislature, and not the Courts, to consider public policy issues and develop the appropriate law.
The Purpose of Tendering
As a starting point, it is useful to revisit the purpose of tendering. The entire tender process is about contracting – the purpose of a tender is for an owner to identify a party who will enter into a contract with the owner. That contract is a private law contract. Once the contract is formed, strangers to the contract have no standing. Tendering is only concerned with the award of an ultimate contract agreed to by the contracting parties.
Elements of Canadian Tender Law
Beginnings – Ron Engineering and Freedom of Contract
It is well known that the Supreme Court of Canada laid the modern, legal foundation for competitive tendering in 1981 in Ron Engineering.
In Ron Engineering, a tenderer sought the return of its $150,000 tender deposit after discovering that its submitted tender, which was the lowest tender submitted, mistakenly omitted $750,058 in costs. Under the terms of the tender documents, any party who wished to submit a tender was required to submit a tender deposit. The tender documents also contained specific provisions under which a tenderer’s tender deposit could be forfeited. The owner could keep the tender deposit if a tenderer withdrew its tender after the submission date or if the owner selected the tenderer’s tender and forwarded the construction contract to the tenderer, but the tenderer did not return the executed contract, required bonds, and other documents within seven days.3
The tenderer submitted a tender as required by the tender documents, including a certified cheque for the tender deposit. When the owner opened the tenders the tenderer learned that its tender was approximately $632,000 lower than the next lowest tender. On review, the tenderer identified that it had made an error in its price. That same day, the tenderer informed the owner in writing that it had mistakenly omitted certain costs, and requested permission to withdraw its tender. The tenderer, however, later changed its position, maintaining that it had not withdrawn its tender. Rather, it argued that, by law, the owner could not accept the tender once the owner received notice of the tenderer’s mistake, thus entitling the tenderer to the return of its deposit. The owner determined that it would not accede to the tenderer’s request and instead elected to send the construction contract to the tenderer for signature. When the selected tenderer failed to return the executed contract within seven days as specified in the tender documents, the owner awarded the contract to the next lowest tenderer and took the position that it was entitled to retain the first selected tenderer’s tender deposit.4
The trial judge agreed with the owner after finding, among other things, that the tender was “filed [. . .] as intended to be filed” and that there was no “error on the face of the tender”.5 The Court of Appeal for Ontario reversed on the basis that an owner could not accept a tender or refuse to return a tender deposit when a mistake in the tender has been “proven by the production of reasonable evidence”6 at some date after tender submission.
The Supreme Court of Canada disagreed, reasoning that the law of mistake did not give the tenderer relief in this context because at the critical time – the time of tender submission – the parties were not mistaken in their intentions and the mistake was not apparent on the face of the tender.
In analyzing the two parties’ rights with respect to the disputed tender deposit, the Court set out its now famous “Contract A” / “Contract B” analysis. The Court started with the proposition that participation in the tender process was purely voluntary:
Here the call for tenders created no obligation in the respondent or in anyone else in or out of the construction world.7
It is not until a tenderer responds to a call of tenders that duties and obligations arise, based on the law of contract, and subject to the terms and conditions in the tender documents:
When a member of the construction industry responds to the call for tenders, as the respondent has done here, that response takes the form of the submission of a tender, or a bid as it is sometimes called. The significance of the bid in law is that it at once becomes irrevocable if filed in conformity with the terms and conditions under which the call for tenders was made and if such terms so provide. [ . . . ] The principal term of contract A is the irrevocability of the bid, and the corollary term is the obligation in both parties to enter into a contract (contract B) upon the acceptance of the tender. Other terms include the qualified obligations of the owner to accept the lowest tender, and the degree of this obligation is controlled by the terms and conditions established in the call for tenders.8 (emphasis added)
The Court emphasized that the parties were in agreement that the tenderer submitted its tender in compliance with the terms and conditions of the tender documents. Further, there was no indication at the time of submission that the tender contained a mistake. Thus, “contract A came into being”, meaning that the “rights of the parties fall to be decided according to the tender arrangements”.9 Under those terms, the tender deposit was “recoverable [. . .] under certain conditions, none of which were met” and also “subject to forfeiture under another term of the contract, the provisions of which [. . .] have been met.”10 Accordingly, the owner was entitled to keep the tender deposit under the terms of Contract A.
The Court observed that “[m]uch argument was undertaken in this Court on the bearing of the law of mistake on the outcome of this appeal.”11 However, the Court concluded that the law of mistake was inapplicable where “[o]n the facts [. . .] no mistake existed which impeded or affected the coming into being of contract A.”12 In thus holding the parties to the terms of their bargain – that is, the terms of the tender documents – the Court agreed with the Ontario Court of Appeal that the “integrity of the bidding system must be protected where under the law of contracts it is possible so to do.”13 (emphasis added) This is the only reference by the Supreme Court of Canada in Ron Engineering to the concept of the integrity of the tender process, and it is tied directly to the concept of upholding the terms of the tender documents.
Judicial Trends since Ron Engineering
The implications of Ron Engineering were not felt immediately. Indeed, with one minor exception,14 it was eighteen years before another case made its way to the Supreme Court of Canada, but gradually since Ron Engineering, the amount of litigation in the area of tender law has increased exponentially, and to date another half dozen cases on tender law have been heard by the Supreme Court of Canada.15
The law of tender has, through this period, become increasingly complex and technical such that it is very difficult even for sophisticated business people to participate in tender processes without the close advice of specialist procurement lawyers. Even with such advice, predicting how the Courts would determine a particular case is very difficult, thereby inserting great uncertainty into procurement. The implications for a project of a legal mistake, including payment of potential profit to an unsuccessful tenderer and risk of delay, can be catastrophic.
Issues since Ron Engineering
“Protection of the Integrity of the Bidding System”
Since Ron Engineering, one theme that seems to have emerged is the need to protect the “integrity of the bidding system”.16 The suggestion seems to be that all tender processes should be measured against something akin to a baseline or intrinsic procurement standard perhaps based on a concern for the “public interest”. However, in Ron Engineering, the Court was speaking to the specific process defined in the applicable tender documents of that case when it made its often‐quoted statement regarding “the integrity of the bidding system” – that is, the system and process that the parties agreed to use when the tenderers responded to the call for tenders. To our knowledge, there is no standard set of requirements of general application to procurements in Canadian common law.
In our submission, since there is generally no common law restriction on an owner in defining the procurement process the owner wishes to use, the only relevant “tender process” in any tender is the tender process as defined in the applicable tender documents. Significantly, in our submission, a Court should resist any invitation to imply terms to, or make interpretations of tender documents which have the effect of imposing additional duties on the parties to the procurement in an effort to make the procurement process meet some objective or ideal procurement standard that does not exist. It must be remembered that each tender process is purely voluntary. The owner offers the opportunity to potential tenderers on the basis of the terms as defined in the tender documents, usually in relation to the terms of the proposed Contract B. There is no obligation on any potential tenderer to submit a tender. Although an owner assumes obligations to tenderers under Contract A, it does not by electing to proceed to contract by way of tender assume legal duties to the tendering community at large. If the terms of the Contract A (or Contract B) are not to the liking of a potential tenderer, that tenderer is free to decline to submit a tender. If no one submits a tender to the owner’s prescribed tender documents, the owner will be required to make appropriate adjustments.
Tendering has not, and the parties involved in tendering per se have not been deemed by legislators in Canada as being in need of statutory protection. There has not been a general movement outside the Courts, apart from limited efforts in the treaty context referenced below, to codify tendering (public or otherwise) processes and impose specific obligations on the parties involved; obligations that would apply in every tender process.
“Duty of Fairness”
To some observers, the result in Ron Engineering may seem harsh, perhaps even unfair. But it nevertheless reflects a fundamental principle of contract law, namely that the parties to a contract are bound by the terms to which they voluntarily agree. The parties to a contract generally do not owe each other a duty of fairness, subject to certain well‐established exceptions such as unconscionability.17
Following Ron Engineering, however, some tenderers have argued that a fairness requirement should be imported into the tender process that is otherwise not found in the law of contract, a requirement that should trump freedom of contract.18 Such arguments, frequently justified as necessary to “protect the integrity of the tender process”, should not, we submit, be accepted by the Courts to the extent such fairness requirements are inconsistent with the terms of the tender documents.
Although fairness does have a place in the law of tendering, its exact role has nevertheless become uncertain through the myriad of Court decisions that have addressed it. To begin, “fairness” generally does not apply to the substance of the terms of Contract A or Contract B. That is, as discussed above, in the common law provinces of Canada, commercial contracting parties generally owe no duty of fairness to each other. In the tendering context, for example, “fairness” does not obligate an owner to make a commercial opportunity “fairly” available to all interested parties. In other words, the fact that one potential tenderer has an inherent advantage (such as previous experience, superior expertise, or a local presence) does not raise an issue of fairness with respect to the law of tender.
An implied term of any tender is that, subject to any express provisions that may be in the tender documents, all tenderers will be treated fairly and equally.19 An owner will have the obligation to treat all tenderers under the specific tender process without bias (either positive or negative) to any tenderer as compared to another; but this duty can be justified under a strict contractual analysis. Thus, for example, an owner will breach its duty of fairness if it evaluates tenderers based on criteria that it has not disclosed in the tender documents.20
In such contexts, however, the duty of fairness should be viewed through the prism of contract law and not a duty outside of contract law. An analysis of whether an owner has evaluated the tenders fairly must necessarily begin with the terms and provisions of the tender documents – that is, the terms to which the parties have agreed (expressly or by inference) at the time of tender close. In essence, this is a matter of basic contract law, including the application of the law setting a high test of the implication of implied terms in commercial contracts.21 No additional duties with regard to fairness are required.
Duty of “Transparency”
Some tender cases, commentators, and certainly practioners, have advocated for the “need for transparency” in tendering. Transparency, it is argued, is necessary to safeguard “fairness” in the tender process. Again based on contract law principles, it is submitted that there is nothing unique to tendering that demands “transparency” and there is no need for the law of tender to impose an additional duty on owners, as part of tender law, in order to achieve some sort of “reverse onus” on an owner to prove that it did not breach its Contract A obligations.
For example, absent a statutory requirement, owners have the right to decide if a particular tender will be open or closed, just like potential tenderers have the right either to participate or not to participate in a particular tender process. Public sector tender documents frequently address the question of transparency by stating that “tenders will be opened in public”, although nothing in the common law and, generally, by statute, requires public‐sector owners to do so. Such notice warns any interested tenderers that their tender price will be public. By inference, this permits a tenderer to either decline to submit a tender, if that potential tenderer wants confidentiality, or to adjust its tender for strategic reasons knowing its tender will be disclosed to its competitors. In any event, the owner obtains the tenderer’s contractual agreement to such disclosure. In many complex procurements, on the other hand, an owner will expressly state the exact opposite – that is that all tenders will be kept confidential.
We submit it is not necessary to impose transparency requirements on tendering so as to ensure fairness given the various legal mechanisms outside of the law of tender by which a participating party may access another party’s tender information in the event of a dispute. For example, the Rules of Court generally provide rights of document discovery by which one party may obtain relevant documents related to a tender. Further, in relation to public bodies, freedom of information legislation grants broad rights of access to information, which may include tender information. Since such mechanisms for accessing tender information exist outside the law of tender, there is no independent need for transparency per se in the law of tender.
In addition, it should be remembered that tenderers are not the only parties in the tender process with fairness concerns. The owner, for example, is interested in knowing that tenderers are not colluding with other tenderers on price. Under the law of contract, however, all parties are entitled to demand that the other parties meet their obligations as defined in the tender documents, but special procedural protections inconsistent with the terms of the tender documents need not be implied to allow parties to ascertain whether the other parties’ obligations are being met.
Public vs. Private Owners?
There is a view that restrictions on an owner’s freedom to contract are particularly appropriate when the owner is a public‐sector owner. There are several different rationales given for this view. One is that because the contract involves the expenditure of public monies, there is a public interest in ensuring that the contract is awarded competitively, at the best price, without the risk of corruption or conflict of interest. Another rationale is that, because the funding is from public monies, then all interested parties have a “right” to be able to participate.
It should first be recalled that there is nothing in the common law that dictates to any owner, public or private, that it must utilize a tender process as the way to identify the contracting party. Therefore, we submit, the mere selection of a tender process should not suddenly impose different duties or obligations on either a private or public‐sector owner with respect to restrictions on freedom to contract.
In our society we certainly need laws to protect taxpayers and citizens against corruption, bidrigging, bribery and other such evils that can occur in public procurement. However, our submission is that the common law of tender is not the appropriate avenue to achieve the required protection. Legislatures have the authority to enact required laws, and indeed have done so. Similarly, there may be public policy or other reasons to restrict or encourage the availability of a contracting opportunity. Government bodies can adopt procurement policies that dictate procurement structures, such as restrictions on sole‐sourcing or rules relating to pre‐qualification. For example, agreements between and among governments such as the Agreement on Internal Trade22 prohibit favouring local tenderes over out‐of‐province tenderers. Similarly, governments’ duties to consult with First Nations can lead to impact benefit agreements in which First Nation contractors will be given preferential treatment.
These, however, are external to the common law of tender. All owners who issue calls for tenders must of course comply with applicable laws that may affect their freedom to contract. A public company may have disclosure requirements that would need to be reflected in tender documents. These external requirements, however, should be irrelevant to an owner’s obligations under the common law of tender, and there is no reason to impose different duties on a public owner as compared to a private sector owner. After all, any owner is simply in a process by which they identify and select a willing party to enter into a private law contract between the owner and that party. That is true whether the owner is a public‐sector owner or a private owner. When signed, Contract B is a private law matter between the contracting parties.
A significant source of litigation in recent years has been the topic of tender compliance. The development of the law in this area has had the perverse effect of introducing more uncertainty and opportunity for mischief. To illustrate this impact, we refer to Graham Industrial Services Ltd. v. Greater Vancouver Water District (“Graham”), a decision of the BC Court of Appeal.23
In Graham, the owner, a municipal water authority, issued a call for tenders for a project to implement various water‐related infrastructure improvements in the Greater Vancouver area. When the tenders were opened, the low tenderer learned that its tender of $21,451,881 was approximately $5,000,000 lower than the three other tenders. It advised the owner that its tender contained a $2,000,000 “arithmetical error” and sought permission to withdraw its tender. The tenderer later changed its position and maintained instead that its tender did not conform to the tender requirements and was, therefore, incapable of being accepted. The owner disagreed and informed the tenderer that it had reviewed the tender and found it to be without any material defects. In making this determination, the owner relied upon the following discretion clause in the tender documents:
10.1 If a Tender contains a defect or fails in some way to comply with the requirements of the Tender Documents, which in the sole discretion of the Corporation is not material, the Corporation may waive the defect and accept the Tender.
The owner subsequently informed the tenderer that it had accepted its tender and forwarded the construction contract to the tenderer. The tenderer, however, failed to execute and return the construction contract within the required time, which entitled the owner to either keep the tenderer’s bid bond or bring an action for the difference between the tenderer’s tender price and the next lowest tender.24
At the BC Supreme Court, the Court sided with the tenderer, finding that its tender was nonconforming and, accordingly, could not be accepted. On appeal, the BC Court of Appeal agreed. The Court determined that the discretion clause could not allow the owner to “subjectively” decide whether a defect was material because the discretion clause only came into effect when Contract A was formed. Thus, the discretion clause could not “give the owner the power to ‘reach back’ before the clause’s operative existence and deem compliant a tender that is, on an objective analysis, materially non‐complaint.”25 In other words, the discretion clause gave the owner the discretion to determine whether a defect or non‐compliance was material “subject to objective scrutiny”.26 Under an objective test, the Court concluded for example that the tenderer’s tender was materially non‐compliant because it failed to provide a detailed, five page outline of its environmental projection plan but rather indicated, in part, that it would “prepare a project specific plan [. . .] following award of the contract.”27
Graham is a case where, like in Ron Engineering, the tenderer selected by the owner, had second thoughts and was seeking to avoid the ultimate contract. This time the tenderer was successful because it persuaded the Court that its own tender, while submitted with the full intention that it was available as the foundation of an award (as was the case in Ron Engineering), was in fact defective so as to make it non‐compliant. The non‐compliancy meant that no Contract A was formed since the tender was not an acceptance of the owner’s offer in the tender documents, and, accordingly, the owner was precluded at law from forcing the tenderer to enter into Contract B.
This legal analysis is well‐founded in contract law principles, but we question the application of that law to the facts of the Graham case. Certainly, where a tenderer submits a tender for a different scope of work than that described in the tender documents, such is a counter offer and it seems clear that no Contract A arises from such a tender. However, that was not the case in Graham.
Interestingly, as a practical matter, the decision appears to offer a tenderer the same opportunities for mischief as were of concern to the Court in Ron Engineering. In Ron Engineering, the Court refused to permit the tenderer to be relieved of its obligations on the basis of a mistake. The mistake was not apparent on the face of the tender submission and the tenderer intended its tender to be the foundation of an award, so the mistake was irrelevant.
In Graham, the equivalent “mistake” was the non‐compliancies, yet as with Ron Engineering, it was clear at the time of tender closing that the tenderer had intended its tender to be the foundation of an award. In light of the law relating to minor or technical non‐compliances, one reasonable view of these non‐compliances is that in the context of this tender process, they were not obviously so serious as to prevent a Contract A from arising, thereby preventing the owner to rely on its discretion clause to waive them as minor. The effect of the decision was to deny the owner the right to exercise its contractually reserved discretion in the tender documents to waive the irregularities in the tender and to relieve the tenderer of the consequences of its “mistake”, in fact making the determination of the materiality of a noncompliance within the exclusive authority of the Court. In a similar case, Manitoba Eastern Star Chalet Inc. v. Dominion Construction Co.,28 the Manitoba Court of Queen's Bench reached the opposite conclusion, on the basis that the tenderer clearly intended to submit the tender it did, and should not be entitled to avoid its Contract A obligations through a post‐closing exercise of combing its own tender for non‐compliancies in order to avoid the result flowing from the Ron Engineering case.
One practical effect of the Graham decision is that in the face of a potential or claimed noncompliance, even by the tenderer itself, to avoid unacceptable risk, an owner may have no option but to either suspend the procurement and apply to Court for an interpretation, or terminate the procurement and re‐tender. The cost implications for the project are significant. Further, this result encourages the process of tenderers conducting a meticulous post‐closing review of their tenders to identify “errors”, or worse, the deliberate insertion of errors to give the tenderer the option to avoid its Contract A obligations, while depriving the owner of its ability to waive such errors under a discretion clause in the tender documents. One view of this scenario is that the owner is punished, while the tenderer is rewarded for its own “mistake”, which the Supreme Court of Canada declined to do in Ron Engineering. A more neutral way of stating it is that, the Courts appear to be straying from the principle in Ron Engineering that where it is possible under contract law to uphold the bargain the parties struck, the Courts should strive to do so.
The Future of Tender Law in Canada
Current State Overall
At present Canadian tender law is very complex and technical, and poses a material risk to parties who participate in tenders. The risks are particularly high for owners, but there are risks for all parties. One reason it is risky is because few construction practitioners fully understand the parties’ rights and obligations. Violation of the technical legal rules is the norm and not the exception in the commercial world. Many of these rules are ones which the parties to a tender could not have foreseen. The result is a legal regime in Canada that is extremely difficult for participants to follow or understand. While the law is meant to give participants in the tender process stability and certainty, in reality it is arguable that we now have a system where only those who are specialists in this area of law can say what is “legal” and what is not. And even then, experienced procurement practitioners are unable to give definitive opinions on the state of the law or how it will be applied by a particular Court to a particular fact situation. Since Ron Engineering in 1981 there has been a significant increase in the amount of procurement disputes so that at some points it would seem that some litigation lawyers had tender disputes as a material part of their practice.
By way of example, the steps taken by the Ministry of Highways and its consultants in the Tercon case were as a result of a failure to appreciate technical procurement requirements. The contract was to construct a difficult highway through a rock canyon to the mouth of the Nass River. Funds were limited and the risk of cost‐overrun was of great concern to everyone. The procurement had changed from a design‐build to a partnering model in which cost savings through on‐site optimization were invited. In that context, the Ministry and its staff welcomed the addition of a new contractor that would work with one of the short‐listed proponents. The tender documents permitted any shorted‐listed proponent to make changes to its team, with the approval of the owner, but such approval was required before tenders closed. In the end this procedure was not followed and on a challenge from Tercon, an unsuccessful tenderer, the trial Court found that the Ministry and its consultant acted in breach of the tender document, and that this breach was so severe that the waiver clause that was included in the tender documents was of no effect.
The outcome of the case was not easy to predict. The BC Court of Appeal and a strong minority of the Supreme Court of Canada did not agree with the result. In the end, however, the Ministry was required to pay Tercon damages in the amount of lost profits Tercon claimed it would have made had it been awarded Contract B.
It seems doubtful that the Ministry and its consultants set out to treat Tercon unfairly or otherwise do anything improper. They were focused on awarding a Contract B (the highway construction contract) for an extremely challenging project. That, of course, was not relevant, but the case highlights the legal risks for an owner in complex procurements.
The technical complexity of Canadian tender law is made worse in areas such as public infrastructure, where consultants and contractors come from other jurisdictions that have different procurement rules. Many of these foreign parties are operating, including providing consulting services (consulting engineers) and contracting services, in ignorance of our Canadian rules.
New Contract Forms
At the same time as we are seeing the technical requirements of tender law increasing, there is a growing demand in the market for procurement flexibility and innovation because of developments in the structure and form of the ultimate contracts. Increasingly, particularly with respect to larger projects, owners, contractors and consultants see the advantages of early (from early stages of project planning and design) contracting to achieve better overall results. We see preliminary contracts awarded to construction contractors for “pre‐construction services”. We see evolution of contract models, such as design‐build, P3, construction management and construction management at risk, in which function planning, design, construction, implementation, operation, maintenance and repair are combined in new ways to achieve holistic benefits over the life of a facility.
These non‐traditional contracts make new demands on procurement because they often do not fit nicely within some of the rules of tender law. For example, early in the development and use of design‐build projects, it was identified that the interests of both owners and tenderers are better served if expanded opportunity is given to tenderers to ask questions, and obtain some level of owner feedback, during the tender preparation stage, prior to the submission of tenders. Such discussions were unthinkable in traditional design‐bid‐build procurements, for fear of tenderers not being treated fairly and given precisely the same information.
These are good commercial developments because they offer an owner (both public and private) the opportunity to achieve optimizations that may not be available in traditional contracting and tendering models. The law of tender, as it is currently moving, has the potential to serve as an impediment to these needs and developments. We urge a movement back to basic contractual principles, including freedom of contract, so that industry can develop mechanisms that fit its commercial needs unencumbered by the uncertainty and costs associated with excessive imposition of uncertain extra‐contractual obligations.