6.3.3  Alliance approach - non-price and price competition processes

Historically, the majority of alliance selection processes have selected proponents on the basis of non-price criteria. This is the method recommended in the DTF Project Alliancing Practitioners' Guide51.

The alternative approach is to select proponents through a process where a significant competitive criterion involves developing the TOC by two (or more) parties. There are also hybrid approaches between the non-price competition and price competition as discussed elsewhere in this Study.

In all Australian jurisdictions there is an emphasis on the integrity of competitive tendering to secure value for money and deliver policy objectives9. Both price and non-price factors tend to be emphasised.

For example, in the NSW Government's Procurement System for Construction83 and its series of Procurement Practice Guides, the Tendering Construction Works guide of July, 2008 states:

A tender evaluation committee will assess each tender against the requirements stated in the RFT documents  The Department of Commerce seeks the best value for money in all its procurement activities for clients. The assessment of tenders is based on all of the evaluation criteria. Contracts and consultant engagements are awarded to the service provider who offers the best value for money, not necessarily to the lowest price.

Although for less complex procurement, price can remain a key determinant. The NSW Procurement Practice Guide Tender Planning for Project Management Services engagements' also of July, 2008 states (page 2):

If the engagement is straightforward and most aspects are well defined, and tenderers are expected to have similar capability, experience and expertise, it may be appropriate to use price criteria only.

The application of such price and non-price criteria, is that competition should be designed so that "the procurement process should optimise effective competitive tension to maximise value-for money opportunities for government"103. This is further elaborated as "Value for money is optimised in the "best bid offer" as assessed against the tender evaluation criteria"50.

The Queensland Government has also published 'Construction Tender Evaluation: Best Value Not Lowest Price' which states:

Achieving value for money typically involves comparing alternatives for the supply of goods and services to get the best mix of quality and effectiveness for the lowest cost over the required term. Importantly, it involves an appropriate allocation of risk, making the selection of a suitable procurement strategy and contract a critical factor in determining whether value for money is achieved.45

In other words, the basis of selecting between competing bidders in public procurement is not the sole criterion of price, rather on selecting the tenderer that on balance best demonstrates achieving the required benefits at the lowest price.

In its Guidelines for managing risks in direct negotiations, published in May 2006, the NSW Independent Commission Against Corruption (ICAC) concluded that77:

Before signing a contract with the proponent, the agency should satisfy itself that, in the absence of competitive bidding, the price paid by or to the proponent is consistent with market values...

...the Commission believes alliance contracting can be beneficial for certain projects, provided the risks are properly managed. Some of the probity risks associated with alliance contracting include:

-  reliance on a non-adversarial approach to conflict resolution and a 'best-for project' approach, which may lead to the parties forming too close a relationship. This may in turn lead to 'capture' by the private sector proponent/s and a failure to consider the overall public interest. Capture can also be a problem if the 'partnership' is lopsided to the extent that the agency develops a dependence on the proponent/s for information and advice.

-  negotiating the target cost of the project after the preferred proponent has been selected, which may remove or dilute the competitive tension that would be present under normal bidding conditions. In this scenario, the successful proponent may have the opportunity to increase the agreed cost or over-design the project.

The above reflects what the Research Team considers to be the cornerstone of good procurement in government. That is, procurement should involve a significant element of competition on outturn price to demonstrate good stewardship of public funds and to optimise both the price and non-price aspects of VfM.

Price is the value placed on what is exchanged. This value includes tangible and intangible factors. Price represents that value and allows buyers to make a choice amongst potential purchases and provides a mechanism for competition amongst sellers in an open market economy.

For comparable products and services, price-only competition requires a seller to constantly innovate and adapt to beat the price of a competitor, and therefore be the lower cost producer. Pure non-price competition requires a seller to compete on only the basis of product features, service, quality etc. Competition based on VfM criteria, which is the common procurement policy criteria for governments, is only possible when there is competition on both price and non price elements. In this later case, buyers assess differences in the products offered and make a choice on the basis of how the non-price differences balance with the price differences.

In the case of alliancing, introducing price-competition as a selection criterion allows the buyer (or Owner) to assess the seller's cost basis compared to its competitors. The buyer will also consider non-price elements to make an informed decision to optimise the VfM outcome. It is difficult for a buyer to make a value for money assessment without considering price as a key element as they need to understand the various trade-offs between price and non-price. Introducing price as a selection criterion provides a positive tension that causes sellers to innovate and provide the best cost solution to address the overall project objective.

However, it is recognised that there are situations where full price competition may not always provide best VfM to the state. Reasons could include high degrees of project uncertainty that preclude project definition or time constraints or there is only one supplier for the required project, product or service. In those situations public officials are normally required to present their reasons and seek approval from senior management and/or Minister before adopting any procurement process that precludes price as a selection criterion.

In other words the default policy position should be to include a significant element of price competition and to depart from this policy must be justified and approved.

In contrast to this view, and despite the emphasis on both price and non-price factors in various general procurement guidelines mentioned above, the DTF Alliance Practitioners' Guidelines expressly states that non-price competition is the recommended approach for alliancing. This suggests that non-price competition should be the default position and any departures must be justified.

It would appear that the focus on price as the primary selection criterion has been in decline relative to non-price factors during the past 10 years or so. For example, the abstract of a paper in the Journal of Construction Procurement in May 2000 highlights the negative perception that has been evident:

The construction industry is dominated by procurement methods that encourage short-term competitive behaviour, driven by price competition. The Australian Procurement and Construction Council has been seeking alternatives in procurement methods, designed to achieve breakthrough change in the Australian industry...73

Moreover, the majority of alliance selection processes during this period have selected proponents on the basis of solely non-price criteria as per the method recommended in the DTF Project Alliancing Practitioners' Guide51.

Some of the main arguments against price competition51, 81 are outlined below along with counter arguments, including observations arising from the Study.

1.  "The alliance relationship is established on the basis of trust and price competition will fundamentally alter this relationship for the worse limiting possibilities and the potential of the alliance."

This Study found no evidence to support the above argument.

A foundation of the alliance delivery approach is that the parties to the alliance take a 'best for project' perspective in determining how to deliver the project and meet the business case objectives. The key requirement in achieving effective best for project decisions is trust between the alliance participants including NOPs and the Owner. There is no evidence that using price competition as a criterion for selecting the NOPs erodes the trust the alliance team develops as the project progresses.

It is important to note that a high performance team can be characterised by the effectiveness of its decision making, and this does not preclude vigorous debate prior to reaching a decision. In fact a lack of competitive tension may lead to poor quality decision making through the effects of group think or misinterpretation that 'trust' means 'no disagreements'.

Owners commented that they did not see any material differences between the effectiveness of alliance teams when selected using either price or solely non-price approach.

The NSW Independent Commission Against Corruption noted that some of the probity risks associated with alliance contracting include reliance on a non-adversarial approach to conflict resolution and a 'best for project' approach, which may lead to the parties forming too close a relationship. This may in turn lead to 'capture' by the private sector proponent/s and a failure to consider the overall public interest. Capture can also be a problem if the 'partnership' is lopsided to the extent that the agency develops a dependence on the proponent/s for information and advice.

2.  "The costs of paying two proponents to develop a TOC (together with Owner costs) cannot be justified."

Noting that the number of price competition selection approaches examined in this Study was limited, the research conducted during the Study indicated that the cost to establish a TOC using price competition was less (of the order of 2% of TOC) than when non- price competition was used. (This recognises that two proponents are paid under price competition.) This saving can be generated through:

•  Better upfront project definition.

•  Fixed timeframe for TOC development.

•  The ability to put a cap on reimbursement of costs for TOC development that is equitable for both proponents.

Furthermore, the TOC established using price competition has been found to be in the order of 5-10% (of TOC) less than when using non-price competition based upon the following elements (which are directly comparable from one project to another) each being higher when using non-price competition:

•  On-site overheads.

•  NOP profit margins.

•  Design costs.

•  TOC development costs.

It was also found that, regardless of the process to select NOPs, the project price and non- price outcomes were viewed favourably by Owner representatives. No disputes between Owner and NOPs were evident that were not resolved within the alliance.

The price competition processes observed in this Study involved reimbursement of costs to the NOPs by either a predetermined lump sum cap or full reimbursement of costs incurred. The Research Team supports this approach of reimbursing the costs to NOPs of preparing their proposals:

•  The proposals involved significant design development costs often beyond that required for a traditional tender.

•  The intellectual property of the unsuccessful proposal resided with the Owner, (on payment of costs) for possible use by the successful alliance team to the Owner's commercial advantage.

•  Non-payment can create entry barriers for mid-tier proponents.

3.  "Price competition limits innovation. Probity considerations reduce the intimate collaboration with the Owner during TCE development which limits ability to innovate."

This Study found little evidence to support this argument. Intimacy is not a pre-requisite for innovation nor does probity limit access to the Owner. There is no apparent reason why formal Probity protocols need to exclude access to the Owner. Rather it is in place, amongst other things, to ensure no proponent is given advantage to information over another proponent, therefore ensuring that equal opportunity exists for all parties to innovate.

The Research Team agrees with the UK Government which has stated that, rather than inhibiting innovation, "vigorous competition strengthens incentives to innovate..." 76

4.  "The pursuit of lowest price could give rise to underpricing risk with potentially unproductive arguments if contingency is inadequate."

This Study found no evidence to support this argument; there was no discernable difference between TOC/AOC outcomes and no evidence of 'unproductive' behaviours.

5.  "There is a significant requirement to control and support two teams in competition together with the need for more external advisors which does not produce net VfM."

This Study found that there is a significant demand on the Owner's management resources. However, there may be a net VfM benefit given the potential for a lower TOC with price competition while non-price objectives are still achieved.

6.  "Price competition is a waste of critical resources - In boom times where design and construction resources are at a premium it is not beneficial to tie up two teams for extended periods."

The statement has merit but the key issue, in the case of any particular proposed alliance, is whether there is likely to be a net improvement in VfM from a whole of government perspective.

7.  "Most non-price competitive alliances have been delivered to ±3% of the TCE (TOC) - The accuracy and VfM of the TCE is verified by the fact that it was very close to the actual outturn costs."

The Study found that there indeed was a high degree of similarity between AOCs and TOCs, with a low average difference and a low overall range of variation between AOCs and TOCs. TOCs were less than the AOCs by 0.5% on average. In addition, the difference between TOC to AOC across the alliances was ±2%.

On face value, these similar TOC-AOC outcomes appear to support the contention within the above statement that an AOC close to the TOC demonstrates VfM. However, closer scrutiny is in order.

Given the inherent forecasting error associated with pre-construction cost estimation techniques, the high level of TOC-AOC similarity is statistically improbable. An estimate is made up of hundreds or thousands of components, many of which cannot be estimated with certainty in advance (e.g. days of constrained work due to wet weather). This inherent uncertainty would typically produce final cost to pre-construction-estimate variations of up to 10-20%, compared to the negligible variation found in this Study.

As a result, the low overall variation range and low average TOC-AOC differences on the examined alliance projects suggests that one or more factors associated with current alliance practices may be distorting outcomes and artificially driving TOC-AOC similarity.

Some possible reasons for the similarity between the AOCs and TOCs include:

(a)  The effort, rigour and therefore cost of developing an alliance TOC is significantly greater than a traditional estimate. (This is unlikely to completely explain the results due to the influence of inherent uncertainty as discussed above).

(b)  The risk pricing basis of the TOC is not consistent amongst Owners and some reflect a higher (more conservative) estimate (e.g. P75) with a commensurate pain/gain share, while other TOCs may reflect a P50 estimate. (This explanation would be statistically unlikely to explain a low variance outcome over a range of projects using different approaches to uncertainty).

(c)  The adjustments to the TOC during the project (found to be 5-10%) are grounded in the emerging knowledge of forecast actual outturn costs, i.e. the adjusted TOC tends to become the AOC.

(d)  The ability of alliance teams to continuously and effectively pursue and deliver savings is not consistent across all alliances and forecast savings that may emerge early in the alliance delivery can tend to be eroded.

(e)  Significant savings to the TOC are perceived by alliances as detrimental to demonstrating VfM and additional scope is undertaken by the alliance without adjustment to the TOC.

The Research Team's judgement is that the most likely reasons are found in items (c), (d) and (e). However, a definitive conclusion could not be drawn from this Study and this issue may warrant further research and investigation. If further research was able to confirm the existence of AOC distorting factors, it could be concluded that AOC-TOC similarity is not an indicator of VfM since the TOC is conservative.

8.  "The TCE is subject to rigorous scrutiny by an Independent Estimator and the commercial alignment framework ensures that the Owner typically receives 50% of the gainshare should it be overstated."

Scrutinising the target cost of the project after the preferred proponent has been selected may remove or dilute the competitive tension that would be present under normal bidding conditions and which would reduce the potential to increase the agreed cost or over-design the project as mentioned by the NSW Independent Commission Against Corruption (ICAC).

The Study indicates that gainshare is negligible and, as previously noted, the TOCs may reflect a level of conservatism much higher than P50 estimates, however, this observation is inconclusive.

Refer to Section 6.5 for comments on the effectiveness of the Independent Estimator.

The choice between a competitive selection process involving only non-price factors and a process involving a price factor should also be considered from an economic efficiency perspective. In particular, the need to optimise:

•  productive efficiency; and/or

•  allocative efficiency; and/or

•  dynamic efficiency.

From an economic efficiency perspective, the Research Team considers that an alliance selection process including competition on price factors is likely to lead to lower costs of project delivery (i.e. maximise productive efficiency), to encourage the allocation of scarce human, capital and equipment resources to highest value uses (allocative efficiency) and/ or to encourage efficient innovation and investment in innovation over time (dynamic efficiency). All other factors being equal, a company competing to become a NOP on an alliance which is optimally efficient in one or more of these efficiency areas at the relevant point in time should be more likely to be selected to participate in the alliance compared with a company which is less efficient.

Given that firms should and probably do primarily seek to maximise profit, price competition would tend to maximise the incentives and focus the profit maximising efforts of industry players on improving their dynamic and/or allocative and/or productive efficiency performance so as to gain a profitable advantage over actual or potential competitors (e.g. to become better at technical or managerial innovation, allocating/deploying scarce equipment or management time etc so as to win jobs).

An approach involving governments signalling in advance (and following through in practice) that they will almost always seek price competition before they agree a TOC is more likely to maximise each of these efficiency factors over time.

This approach can be consistent with a focus on competition (efficiency) behavioural drivers during the period before the TOC is settled and which focuses on collaborative problem- solving behavioural drivers after the TOC is settled.

In other words, a NOP selection process using price as a key criteria will tend to optimise economic efficiency in the long term from all relevant efficiency perspectives (productive, allocative and dynamic) and hence optimise VfM from a whole of government perspective.