Tie-breakers
Where clients come out on the balance will hinge on commercial considerations such as time pressure, the structure of the market and, especially, whether there is an uncooperative, aggressive, incumbent supplier with a lot to lose.
This means that there is a balance to be struck when advising on risks which are, by their nature, hypothetical. While I prefer to keep it pragmatic, we do sometimes have to consider some purely hypothetical risks, such as: “what happens when there is a tie?”
1. When are bidders “tied”?
There is a fairly well-established practice for a well-run procurement process. The contracting authority will publish a detailed soring table, setting out the evaluation criteria, questions, mark-schemes, scoring criteria and weightings for each element. This is often accompanied by a spreadsheet demonstrating how the scores can be calculated to two decimal places. When a procurement exercise is scored on the basis of multiple questions and multiple weightings followed by, perhaps, a moderation meeting, it makes the chances of a precise dead heat pretty minimal.
On the other hand, it is normal to have bids which are very close. Sometimes very close. I have been involved in procurement challenges where the difference between the winner and next bidder was less than 1%. In those circumstances, it is no surprise that a losing bidder might challenge – a point here or there might make the difference to someone between getting a sales bonus or getting the sack – but on the other hand, if there has to be winner there has to be a loser. But that of course leads us to the question – where the decision, on paper, is so close, and the client is neutral as to the bidders, it might make sense to treat that as a tie.
The contracting authority could treat the gap between the winning and losing bidders, if they are within a tiny range of each other, as not being “statistically significant”. This could be the cumulative effect of rounding up scores which leads to a margin of less than 1% in my earlier example. So this leads us to a question as to whether it is sensible to treat two scores as being so statistically insignificant as to be treated as a tie? This is not the orthodox position, and you can see immediately how this ties the contracting authority into knots: if a bidder has won on the basis of the scored arithmetic but then is not awarded the contract, then this is asking for a challenge.
2. What to do when bids are tied?
Procurement law treats tie-break criteria in the same way as it would for any other evaluation criteria. The European Court of Justice dismissed the attempt by the French authorities to treat a tie-break criterion as a “secondary” criterion, which is subject to a different, lower level of scrutiny, in the event of a tie-break on the main evaluation criteria. But the use of tie-break criteria becomes, at the stage that comes into play, the decisive criteria for the contract award, and so should be treated as award criteria.
It is instructive to look at approaches from the UK and elsewhere. Guidance which is published by a well-known publisher in the legal industry suggests either of the following suggestions:
“- Setting additional questions to be answered.
– Re-opening certain parts of the tender to be re-evaluated in writing or through a presentation.”
While it is certainly useful to set out the process, it still seems to fall a little short of the general procurement principles if the rules of the tie-breaker do not set out the questions to be answered or the parts of the tender to be re-evaluated. Which leads to the point that if you identify those questions or those areas, then why not include and score those questions in the first place; and if you can score those questions, why not take the results of those questions or areas as the tie-breaker? This has the merit of reflecting the contracting authority’s priorities, and is used, for instance, by the Department for Work and Pensions as follows:
“…the Authority will apply a tie-breaker in the event that two or more Bidders have the same combined score. Where a tie-breaker is required, it will be applied to the scores of all Suppliers with joint scores.
11.16 The application of the tie-breaker will be as follows. If at any step this produces a clear result which differentiates the Suppliers who have the same score, the process will be terminated.
Step 1 Highest Score Delivery Proposal and Target Audience 10.3
Step 2 Highest Score Programme Outcomes 10.5(a)
Step 3 Highest Score Participant Journey 10.4 (a)
Step 4 Highest Score Participant Journey 10.4 (b)”
This same approach is used by other public sector clients, using other award criteria reflecting their relative priorities:
“In the event of a tie break or statistical tie break (i.e. where one or more of the highest scoring Tenderers has an overall score that is within 0.5% difference), the Council will award the Contract to the Tenderer offering the lowest price.”
Similarly, previous guidance from the Office of Government Commerce, Making Equality Count, tentatively suggested the use of “equality” as a tie-break criterion where other scores are equal, reflecting how much equality counts.
On the other hand an OECD report on the public procurement regime in Mexico indicates that in the event of a tie, micro, small and medium sized enterprises should be favoured, in that order. Failing that the authority will draw lots to pick the winner at random, thereby reducing a detailed and complex process accessible only to spreadsheet enthusiasts to a one-off luck of the draw – simple, easy to understand and apply. I understand the ICC is even now looking at Mexican procurement law to see what lessons can be learned.