Journal Issue
Finance & Development, June 1978

Insuring major civil works: After surveying the better-known coverages of major construction projects, the author suggests alternatives that could eliminate wastage and overlapping insurance protection

International Monetary Fund. External Relations Dept.
Published Date:
June 1978
  • ShareShare
Show Summary Details

David M. Sassoon

The execution of a civil works project normally entails the purchase of several types of insurance by the owner, contractor, and consulting engineer, respectively. There may be as many as five different policies for the construction period alone, and many others for the post construction or operational period. The type of coverage is obviously determined by the nature, complexity, and location of the individual project, but the construction of a major civil works project will be normally insured for at least: (1) Contractor’s All Risks (CAR); (2) third party liability; (3) employer’s liability; (4) constructional plant and motor; and (5) professional indemnity or liability. If parts of the work are subcontracted, further policies will be required. Some projects involve up to 50 separate covers, and two dozen are not uncommon. In addition, the owner, usually a government or some other public entity, normally requires that all the firms participating in the construction of the project post a performance bond or bank guarantee to provide some protection against the financial failure or noncompliance of any of these firms.

How far do the traditional bonding and insurance patterns protect the owner? What are the alternatives? Are there ways to reduce costs without significantly impairing protection? This article examines the implications of these questions. The solutions proposed are particularly relevant for projects where the procurement process follows established rules, and the source of financing is one (such as an international lending agency) that is capable of exerting some influence on all the parties involved.

Performance bonds

It is generally assumed that performance bonds (or bank guarantees) are indispensible. First, they safeguard the owner against the risk of the contractor abandoning a project if it proves unprofitable, or if he is incapable of completing it satisfactorily. Second, these bonds increase the chances of completion if the contractor is threatened with insolvency. However, assuming the owner has taken reasonable care in selecting his contractor, the added costs incurred by posting these bonds may well not be justified in each and every case. Taking “reasonable care” would involve the owner selecting an experienced, competent, and financially sound contractor, and signing a contract based on a bid which is not below the real cost of the project—which will almost invariably be the case where the project is being financed by a major and experienced international development lending agency with broad procurement scrutinizing powers. Provided there are no unknown technical factors and the project risks are adequately covered by insurance, the performance bond will then only protect the owner in the event the contractor was, for reasons entirely unconnected with the specific project, threatened with insolvency. Although such situations have arisen, their number, in the case of large projects involving major international contractors, is believed to be insignificant. The total cost of performance bonding of major civil works contracts probably far exceeds the amounts ever collected from calls on them.

Thus, if a project contains no exceptional features and the financial position of the contractor is not in doubt, the owner may well decide not to call for a performance guarantee. However, some international lending agencies still require the owner to obtain such security in the case of all civil works contracts financed by them on the basis of international competitive bidding procedures (see para. 2.14 of Guidelines for Procurement Under World Bank Loans and IDA Credits). In these cases, there may be a more rational and less expensive way of spreading risks between the various beneficiaries of loans from international lending agencies.

Alternatively, it may be possible both to take measures to reduce the exposure of the owner to the risk of financial failure by the contractor, and also effect certain savings by reducing the value of the bond and/or guarantee. The Overseas Private Investment Corporation, which insures U. S. contractors against political risks, has recently managed to convince several Middle Eastern countries to reduce their bonding or guaranteeing requirements to 5 per cent from the prevailing higher levels. This has involved, inter alia, limiting the owner’s exposure through early transfer of title to equipment and materials, which means that labor and other local costs which are in the control of the owner need not be covered. Moreover, when the owner takes early title to equipment and materials—upon delivery in the project country or on site—he minimizes the risk of the equipment being seized, and of interference in project execution if the contractor either becomes insolvent or abandons the project.

CAR insurance

Contractor’s All Risks (CAR) insurance is generally used as the main protection against physical loss, damage, or injury to the project. CAR insurance is normally issued in the joint names of the owner and the contractor, and often includes excess or deductible provisions as well as ceilings on the liability of the insurer. This is particularly so when the project is very large and the insurer wants to set a limit to his total liability (or to his liability in respect of individual claims) in advance.

Owners (particularly governments or other public authorities undertaking a large number of civil works projects) would be well advised to obtain expert advice in order to scrutinize carefully the policy offered by insurers under this policy so as to maximize coverage and save premium costs. Lenders such as the World Bank normally require the owner to obtain adequate insurance coverage in accordance with customary practice. This is not designed to relieve the owner of this task. Lending institutions will, as a rule, accept practically any customary insurance coverage as satisfying their requirements.

The CAR policy often includes elements of both over- and underinsurance from the owner’s point of view. Although the contractor should be bound to insure only those matters which are his responsibility under the contract, the standard policy normally contains elements which may be of little or no interest to the owner. For example, insurance against minor or partial damage to the works for which the contractor is liable is often redundant from the owner’s point of view. Certainly, it is in the owner’s interest (though apparently seldom realized) to have a large excess in respect of these items. Otherwise, the wide ranging coverage here may act as an inducement for contractors to cut corners and minimize precautionary and protective measures in order to maximize profits. On the other hand, the cost of repair or reconstruction of defective work or materials which is normally excluded from the insurance may be vitally important to the owner where major civil works are concerned, if the failure occurs at a late stage of the construction.

Faulty design

There is one important standard exclusion in the CAR policy which is likely to be invoked by insurers in a significant number of instances of major loss or damage during construction. This exclusion concerns “loss or damage arising from faulty design and liabilities resulting therefrom” or words to that effect. It denies indemnification for the cost of making good the faultily designed part, and, depending on the language of the particular policy, may also exclude protection for any loss or damage arising as a consequence. The first observation on this exclusion is to query its purpose. Why postulate such a provision if the CAR policy is only designed to cover those causes for which the contractor assumes liability? Loss due to faulty design is obviously not such a cause. Surely, if the purpose is actually to deny coverage for loss or damage for which the contractor is in fact liable, then the practice is questionable indeed because this is precisely the type of insurance protection the owner expects and presumably assumes he is getting from the CAR policy. One must therefore assume (or hope) that the faulty design exclusion adds nothing to (and is no broader than) the immunity of the contractor under the concept of “faulty design” (for which the contractor would normally not assume responsibility). If the purpose of this exclusion is to alter to some degree the legal position between owner and contractor, it is a very serious matter indeed. One argument that lends support to the claim that the exclusion does indeed alter the legal position between the parties is that it is usually couched in language that is different (and on its face somewhat broader) from the language found in the underlying construction contract. This often provides that the contractor shall not be responsible for loss or damage from a cause solely due to the engineer’s design of the works.

Another point concerning the “faulty design” exclusion in the CAR policy is the following: Does any element of personal failure or noncompliance with the standards of the engineering profession have to be present in order to exempt the insurer from liability? In Queensland Government Railways and Electric Power Transmission Pty. Ltd. v. Manufacturers Mutual Insurance Ltd., decided by the Australia High Court in 1968, this question was answered in the negative. In that case prismatic piers designed to support a railway bridge were overturned by flood waters after exceptionally heavy rains. The CAR insurers denied liability and the Court held they were entitled to do so, because no element of personal failure by the design engineer was required for the insurer to escape liability under this exception. It was enough that the design did not work. In other words, the decision held that the exclusion was not against loss from “negligent designing” but against loss from “faulty design,” a term which is more comprehensive than the former. The construction contract usually contains no legal obligation on behalf of the contractor to advise the consulting engineer (or owner) of any design defects he may suspect or detect in the course of performing his contract.

In addition to the “faulty design” exception, which is significant, insurers have sometimes argued that many losses in major construction projects were not fortuitous in origin. That in fact they were inevitable—although this may not have been known at the time—because knowledge in the field was not sufficiently advanced. Where the insurer is able to show that the proximate cause of the loss was not a “risk” but a certainty, he will not be liable to provide indemnity. It is a basic principle of insurance law that the insurer is liable only if loss can be attributed to a risk or an accident, and, except in the case of life insurance, the policy does not provide protection against a certainty, that is, an event that had to happen in the circumstances.

Thus, where the insurer is able to establish the nonfortuitous nature of the cause of the accident, the owner is in fact not covered by the policy although he may reasonably expect that he is. This, of course, ignores for the moment the extreme difficulty of reconstructing with absolute assurance and beyond reasonable doubt the precise interaction of events, factors, and causes of major failures in the case of complex projects such as dam or tunnel cave-ins.

Professional liability

What happens where the CAR “faulty design” exception does apply? The owner will clearly wish to recover from the consulting engineer the loss or damage sustained.

But in order to recover from the consulting engineer under the latter’s professional liability policy, the owner must show a breach of professional duty, or negligence on the part of the consulting engineer. Legally the scope of the consulting engineer’s professional duty is no more than to exercise the ordinary skills of a competent member of his profession.

Consequently, not only can the CAR insurer escape liability on the “faulty design” exception in the CAR policy, but in order to recover under the professional liability policy for loss or damage due to this very same faulty design, the owner must first prove a breach of duty on the part of the consulting engineer. The combined result of these two principles is that the owner may find he has no protection whatsoever under the terms of either policy for loss or damage due to faulty design.

Furthermore, the owner may only find this out after very prolonged and very costly legal proceedings. Not only may the CAR and professional liability policies be subject to different jurisdictional and choice of law clauses (as will almost invariably be the case in international projects) but a finding under one policy will not be binding under the other, even where the proceedings are held in one and the same country and before the same judge or court, because, inter alia, the defendants are different. The CAR insurer and the professional indemnity insurer are seldom if ever identical even in the case of turnkey contracts. Specialization, tradition, and vested interests have created two separate and distinct insurance markets in this area, and the owner must contend with both on an entirely different basis.

Even where the owner is able to establish a breach of duty on behalf of the consulting engineer he may find that the engineer’s resources (including the professional liability policy) are far below the amount of damage suffered, because professional liability insurers are reluctant to underwrite risks above a maximum ceiling presently estimated to be £5-10 million at the most in London. Moreover, professional engineers are increasingly resorting to contractual and other means (such as incorporation with limited liability) of limiting their legal liability. While these measures are normally explained by the need to relieve from the design professional an “unfair” burden for the unexpected occurrence, they in fact limit or avoid a legal responsibility that would have otherwise obtained.

One must not, in the final analysis, lose sight of the fact that in deciding to go ahead with a project the owner has relied on the consulting engineer’s advice that the project was feasible and could in fact be executed safely and economically.

Are there any alternatives to this situation which may increase the owner’s protection and reduce his costs? Clearly, the large number of policies involves some waste of funds and overlapping of the areas insured. There seem to be two alternatives. First, less or no insurance, which means self-insurance for major and complex publicly owned projects; cost/ benefit analysis may indicate this to be a sensible course of action in light of the past loss claim experience in the area. Second, a comprehensive project insurance approach under which cover would be obtained by the owner himself instead of being issued by the several individual project contractors, subcontractors, consultants, and suppliers. Clearly, the owner stands to gain, both financially and otherwise, in a situation where he has to deal with a single insurer over one policy. One may safely assume that this would involve simpler and less costly claim settlements.

Is comprehensive project insurance feasible? Could it also encompass liability for faulty design and make the professional liability policy redundant? There is expert evidence which unequivocally shows that this would be a wise approach to adopt for major civil works projects. As far back as 1968, one of the leading international insurance brokers advocated this approach which he described as Project Insurance. (See E. M. Saventhem, “Introduction to the Insurance of Construction Projects.”) In addition to providing direct benefits to owners, such an insurance would also, in his view, be of benefit to the economies of developing countries because “project insurance offers exceptionally flexible facilities to reserve, for the local insurance market, the highest degree of participation compatible with the overall objective of maximum rationalization and economy.” He rejected the claim that such insurance would be more expensive because it would disregard the experience and personal relationships of individual contractors and their insurers. Saventhem returned to this theme in 1973 when he said: “… In my firm’s experience with large civil engineering projects, we have often been struck by the wastage of the owner’s money which results from the system of making every participant arrange separate insurance policies for his part of the work, his own plant and equipment, and for the liabilities to workmen and to third parties which might arise from performance of his part of the work. We have therefore been advocating comprehensive project insurance to be arranged by the owner. Our approach so far has been confined to conventional risks—damage to the work and to other property at the site, and liability vis-à-vis third parties, including cross-liabilities between participating engineers and/or contractors. There is no reason—at least not in insurance theory—why this concept should not be broadened further to include the consulting engineers’ professional liability.” (See E. M. Saventhem “Professional Insurance,” The Consulting Engineer, Sept. 1973, p. 25.)

Similar views were voiced by at least two other reputable international brokers with whom the author discussed this topic. It was their considered opinion that by adopting a project insurance approach the owner could obtain a much larger coverage for loss due to faulty design than he gets nowadays—indirectly—by means of the professional liability policy of the consulting engineer.

Interestingly enough, such a project insurance approach has recently been adopted in the case of certain public works in Norway, and is also being actively advocated by some of Europe’s leading brokers. The World Bank has not yet had any practical experience with this concept, although it recently recommended to one of its borrowers that this alternative be examined before a final decision was reached on obtaining the required insurance for a major dam the borrower was proposing to construct.

What of the future? Several factors seem to be combining to stimulate a fresh approach to insuring major construction projects. These include certain legal developments concerning the liability of the engineering profession as well as a greater awareness of the risks involved in major civil works schemes generally. The International Federation of Consulting Engineers is planning to sponsor a seminar of experts to review this entire subject in order to explore alternatives to the present regime governing the insurance of major international civil works schemes. Hopefully, the results of this seminar will provide all parties concerned with a better understanding of their options, the costs thereof, and the optimum protection they should seek in the various situations confronting them.

A major issue one should address when considering insuring major civil works projects is whether project insurance, assuming it is feasible, should prevent recourse against a defaulting contractor and/or engineer. The latter prefer something akin to a no-fault approach, but the analogy to automobile insurance which they invoke is doubtful. Criminal law sanctions (against dangerous driving) and the risk of bodily injury to the driver are sufficient deterrents to no-fault being abused by vehicle operators, but no similar safeguards (against cutting corners or sloppy design or execution) exist in the case of civil works schemes.

Other Resources Citing This Publication