Highways Agency

Results Of Discussions With Key Organisations

Introduction

Following the analysis of the responses to the consultation document, a number of individual meetings were held with a broad cross section of the organisations that had provided responses. The purpose of these discussions was to probe the responses and explore the issues raised in more detail. In addition to the face to face meetings, discussions were also held by telephone with a further sample of organisations including some that had not provided a response to the initial consultation document and whose opinions we felt could be valuable.

This section of the report summarises the additional information and insights gained from these interviews. For ease of reference, the results are set out broadly in the order in which the issues were raised in the consultation document. However it was clear from the meetings that many of the issues are inter-linked and do not always neatly fit into the categories set out.

Definition of Performance Specifications

There is an issue of definition that underlies any consideration of performance specifications. The boundary between output and outcome specifications is not always fixed or clear. For example, a specification that requires grass in highway verges to be cut to always be below a height of three inches, rather than a specific number of times in the season, might be regarded as an outcome specification. However a higher level outcome specification might be to require grass to be cut to a standard which is consistent with the safety and environmental requirements of the contract. If the latter were also specified at the outcome level, there would be considerable scope for interpretation and innovation by providers with potential benefits in terms of effectiveness, but also with an inherent risk that the provider's interpretation of acceptable standards may not coincide with the client's.

An example of such a difference of interpretation where the actions taken by the provider might have potentially undesirable consequences would be for the provider to seek to reduce traffic delays caused by roadworks on part of his network by diverting traffic to local authority roads. Another example might be the use of concrete barriers instead of safety fencing on central reservations to reduce cost.

A further distinction which became evident is that whilst measurement of outcomes was acceptable within contracts, possibly as a basis for benchmarking; payment based on outcomes was far less so, particularly if there was a risk that costs incurred by the provider would not be met if outcomes are not achieved.

Benefits of Performance Specifications

Respondents generally felt that there were real and worthwhile benefits in working under a performance specification and more particularly under an outcome specification. These benefits include flexibility, innovation and better integration of design and construction processes.

As one respondent put it, a performance specification 'allows the resourcefulness, experience, imagination and ingenuity of the design and build team to be fully realised'.

A number of examples were given of measures or actions that providers might consider introducing under performance regimes:

• More use of traffic modelling to improve the design of roadworks so as to minimise traffic delays.
• If gully cleaning was subject to an outcome measure, contractors may target gullies prone to flooding with more frequent maintenance and reduce cleaning frequency where not critical.
• Emergency response using motorbikes (trialled in one area)
• Possible use of vehicle-actuated speed warning signs to encourage driver awareness of speeding and improve road safety
• More pro-active approach to customer information, for example keeping databases of people who have commented or complained in the past and providing them (if requested) with better information
• Road markings could be specified on an output basis with the contractor responsible for maintaining them at a specified standard for a number of years
• More careful consideration of maintenance options, an example was given of a failing concrete culvert which was successfully dealt with through a regime of repairs and inspection which the provider considered to have been more cost-effective than reconstruction

However, the impression gained from the interviews was that most providers have not as yet developed clear and well thought out strategies for re-engineering their approach or their processes under an outcome specification based contract. In particular, the possibility of a whole life costing approach leading to significant changes in the way providers deliver the service did not feature significantly in the responses.

It is likely that providers have as yet not fully considered their commercial response to a different (possibly radically different) contract basis and may not do so properly until they see the details of any proposed new contract forms. An additional factor is that many respondents did not support a full outcome based contract (the option 3 scenario) and felt that this is unlikely to be introduced for some time.

Elements Suitable for Performance Specification Contracts

Generally it was felt that performance specifications were more suitable for routine maintenance activities and more workable for long term contracts of 20 years or more duration. Most providers are as yet unconvinced of the feasibility of some of the high level measures put forward in terms of their suitability for use as a basis for payment.

The provider's ability to determine changes in traffic accident levels was particularly questioned and there were also concerns about providers' ability to influence journey time reliability and customer satisfaction. However this view may change as prototyping of these measures proceeds and providers gain more confidence in the measures proposed and in their ability to influence outcomes.

There was also a widely held view that the long term durability of pavements was difficult to predict and was not a suitable element for performance measurement. This was partly because of the limitations of existing knowledge of the performance of pavement materials and structures under different site conditions and partly a concern about the effects of external factors which are not under the provider's control, such as the type of traffic using the road.

It was also considered that there is a difficulty in reliably quantifying the residual life of the network at the handback stage. This would need to be overcome in order to enable suitable handback provisions to be included in contracts to ensure that providers do not reduce the level of maintenance towards the end of the contract.

Measuring Outcomes

There was significant support in principle for moving forward with the introduction of outcome measures into Highways Agency contracts, although there was no clear agreement as to which outcomes should reside with the HA and which with the providers. It was noted that where the responsibility for the delivery of outcome measures rests with the providers, they must be achievable, measurable and deliverable. An example of where factors not under the control of the provider could reduce the validity of the measurement of the provider's performance is where congestion in adjacent HA areas caused knock-on congestion on the provider's network. The possible development of the current DBFO contract KPIs was suggested as a way forward with an initial emphasis on outputs rather than outcomes. Other approaches such as the provision of detailed method statements at tender stage, the use of integrated management systems and operation within agreed quality management plans were also mentioned.

Time is needed to understand and develop the measurement process and this will require additional research and development and testing over and above that being undertaken currently by the HA. Many interviewees noted that the majority of organisations involved in HA commissions do not undertake or invest in any major R&D and therefore a process of piloting and testing may be the most appropriate way forward.

Risk Transfer Issues

The assumption of greater risk was approached with caution by most respondents and particularly by consultants who generally felt that the level of risk they were required to carry should be capped. The key message was that providers should only be required to carry risks which they had the ability to manage.

Latent defects were held up as an example of risks which are outside the provider's control and would make a contract based on an asset condition performance specification unworkable. However it was acknowledged that the problem may be less significant where contracts cover a wide network. Other risks mentioned as not suitable for transferring to providers or would impede the introduction of performance contracts included the following:

• changes in client objectives and changes in legislation (eg higher permitted axle weights). A further example given was the possible introduction of motorway tolls which might encourage night time driving and make it more difficult to carry out night time maintenance.
• Third party interaction including:
  • damage to the network
  • traffic generating development schemes affecting network performance
  • disruption of the network (eg because of accidents or statutory undertakers' work)
• Changes in road use, particularly in terms of HGV volumes and weights.
• Numbers of accidents occurring on the network
• Network inventory inaccuracies and omissions
• Funding changes
• Changes in standards
• Outcome based payment mechanism not covering costs.
• Mismatches between contract performance requirements and client objectives. As one contractor put it 'The risks of performance specification is that design and build team builders under specify, which results in poor operational performance, excessive maintenance and premature replacement.'
• Providers' ability to control network performance
• Lack of management or engineering expertise and lack of skilled resources generally within the highways sector

Client respondents expressed concern at the feasibility of transferring risk from a practical legal point of view of demonstrating the provider's responsibility (eg concrete spalling occurring decades after construction) as well as the risk of providers' inability to meet their obligations should the need arise (long term financial viability of commercial enterprises).

The transfer of risk was perceived to be particularly problematic for small organisations which may not have the financial strength to absorb substantial levels of risk. There was a perception that the cost of insuring the risks related to performance specifications would be high, although none of the firms interviewed had as yet tested the insurance market in this respect. By contrast, some materials suppliers were concerned that risks would be passed down the supply chain and that as large companies (in comparison to most contractors) they feel that they 'have more to lose' if inappropriate risks are taken.

A further issue in relation to risk is that providers will look at risk in commercial terms, which may not coincide with the political perception of risk. It would be difficult to frame commercially acceptable performance contracts that realistically reflect political risk.

A number of interviewees stated that providers prefer to have steady assurance of turnover and profit and therefore will not take unnecessary risks for the opportunity of possible large profits. If innovations require high risks, they believed that these should be developed by the HA in partnership with providers to ensure that continuous improvements are assured and risks minimised without detriment to providers' businesses.

There was also a concern that acceptance of risk by providers may lead to more conservative design decisions. Sub contractors and suppliers may not be able to fund their risk allocations or obtain the necessary insurances and insurance companies are likely to require client approval for decisions, new techniques and innovations in order to reduce the risk to the provider and themselves

It was considered that when risks are appropriately allocated and shared where necessary between the HA and the providers, pricing will be more accurate and transparent, providing greater value for money. It was additionally noted that the benefits of partnering and risk sharing would be increased if the HA was sharing both the pain and the gain in an equity or other arrangement.

A number of methods for managing risks appropriately were suggested, some of which are summarised in the following list.

• Effective and thorough selection of providers based on both their own risk management techniques and those of their supply chain
• Early involvement of supply chain in the procurement process
• Partnership principles applied to the management of the whole supply chain
• Supply chain agreements that have clearly defined risk allocation linked to payment
• Retention by the public sector of risks that can not be appropriately allocated within the supply chain or insured against
• Incentivised supply chain arrangements, including back-to-back risk sharing limited to the particular activity that is being subcontracted
• Robust risk mitigation plans

Cultural Change

There was a general view that a change of mindset towards more openness and trust would be required for both the HA and providers under a performance specified contract. Providers would have to be prepared to take on more risk and this might be less easy for consultancy firms than for contractors who are more used to operating in a risk environment.

It will also be necessary for the HA to concentrate on monitoring outcomes and accept a more hands-off role in terms of monitoring detailed specifications. As an example, it was pointed out by one of the providers that Sector Schemes which were intended to lead to less client supervision, by providing some quality assurance, including effective training of operatives, have not yet resulted in a reduction in client supervision.

There is an understandable reluctance within some parts of the HA to move in the direction of reducing direct supervision of processes until there is clear evidence that performance contracts are resulting in providers delivering the required quality and that the measures used are sufficiently robust to ensure the right result.

Technical Governance

Some providers identified potential benefits if the HA were to pass responsibility for standards to suppliers, as illustrated by the following comments: 'The impact on suppliers will be one of improved professionalism' and 'Engineers should be encouraged to engineer solutions instead of the prescriptive "easy option"'

However, the concept of the HA exercising its technical governance role by passing responsibility for standards to providers was generally not supported. The reasons for the lack of support seem partly cultural because the industry has not been used to 'guaranteeing' its products in terms of either durability or fitness for purpose.

A second and related reason is the lack of adequate technical knowledge to enable such 'guarantees' to be responsibly given, this in turn is a reflection of the fact that the industry currently looks to the HA for much of the R & D required. The scale of change required is considerable and is likely to take some time. Some clients and providers questioned whether it would be in the HA's interests to become a less technically involved organisation. These issues have to be weighed up against the benefits which were set out earlier. It was also noted that the HA is in a better position to justify expenditure on R&D given its responsibility for the entire network whereas providers may find it difficult to justify such expenditure on the basis of a single contract. A contractor noted that 'If the responsibility for full technical compliance was passed to the supplier each would have to devise R&D plans along with long term monitoring recording and feedback systems…'

Client staff were generally concerned at the potential loss of quality if only outputs or outcomes were specified with no detailed technical specification. A number of clients pointed to examples of poor compliance with existing detailed specifications as evidence that standards are likely to fall further in the absence of such specifications. However, there was also a view that some of the non-technical requirements of existing specifications (eg the parts of DMRB and MCHW concerned with the management of maintenance programmes) could be removed if robust performance specifications were in place. There was also an HA client view that the existing departures process did not inhibit innovation and perhaps needed to be better communicated to providers. This view was supported by a local authority client who commented 'Desired benefits could be achieved without the perceived risks by maintaining the existing procedure but adopting a more flexible approach to accepting suppliers/contractors proposed alternatives'

Some providers felt that the geometric standards within DMRB should be retained but there may be less necessity for detailed construction related standards.

A further issue mentioned was the potential impact on third party claims if there were no generally agreed technical standards in place.

There is support for more involvement of industry in the setting and approval of standards and a number of suggested ways of achieving this were put forward, including an expanded role for the Roads Board or some form of new technical standards approvals body. All respondents agreed that the HA should continue to be involved in the setting of technical standards at the very least in a coordinating role and a number of both clients and providers believed that the HA should maintain its leadership role in terms of technical expertise. It was noted that 'the Steel Construction Institute is an example where quality has been monitored / improved by a suppliers representative body'.

A number of organisations pointed to the wide use of HA standards such as SHW and DMRB by public and private sector clients outside the HA. Some public sector clients feel that they should be more involved in formulating the standards and a number of providers would favour such a development particularly if it led to nationally agreed standards which can be used on all UK highway projects, as this would reduce their 'learning' costs in dealing with different standards.

Ensuring Quality and Continuous Improvement

Some providers believed that suppliers would develop their own specifications over time and would have to update these in order to remain competitive. One commented that 'With an outcome specification for new contracts there is no need to ensure that specifications and standards are updated. Updating will occur naturally as suppliers seek to maintain competitiveness'.

The need for some form of change mechanism within contracts to enable requirements to be updated in line with changes in technology or for other reasons was referred to in some of the interviews.

Impact on Tendering

A concern expressed by a number of respondents, particularly clients, was that the introduction of performance specifications may lead to a reduction in the number of firms prepared to tender with a risk of increased costs. There is some evidence from local authority clients and from contractors that this has already happened in some instances. One company had experienced having to incur a large loss as a result of agreeing to enter a performance contract for pavement durability (for which the measurement basis was limited to only one type of durability test). The firm commented that they would be very unlikely to bid for such a contract in future as the risks are considered too high.

Providers commented on the high cost and risk associated with tendering and suggested that the HA should consider reimbursing a proportion of tendering costs incurred by unsuccessful bidders. The complexity of tenders was commented on. One company has adopted the practice of deferring detailed due diligence checking of the contract until after success at the bidding stage, this was due to the high cost of such vetting of tenders.

A number of providers emphasised that the availability of robust network data is particularly important in tendering for a performance based contract and providers felt that the quality of asset data would need to be considerably enhanced to enable them to tender for a performance contract. A number of respondents commented that the evaluation of tenders under a performance specified contract may become more complex and less transparent as it would be difficult to compare tenders which are prepared on the basis of differing standards.

Payment Mechanisms

The need to cover costs and secure a reasonable return on investment is a key concern for providers when considering the linking of payment to performance measurement. If too high a proportion of contract payments are linked to performance, potential tenderers may be discouraged from bidding or bid prices may increase significantly to cover the risk.

Some providers expressed a preference for an open book system with basic costs being met and with additional performance payments linked to achieving specific outcomes.