Dutch design competitions : lost in EU directives? : Procurement issues of architect selections in the Netherlands

on public service contracting 2004/ 18/ EG & 2004/ 17/ EG) were introduced, they were associated with bureaucracy and high transaction costs. During the past several years, these frustrations seem to have increased rapidly. In the Netherlands, architects have refused to take part in architect selections, have ended up in court fighting jury rulings and fellow architects, and had the issue addressed in Dutch parliament.

Introduction « A ritual dance » « A beauty parade » « A bureaucratic straight jacket » That is how several of the interview-ees in this research qualified the current architect selections for public commissions in the Netherlands.To a certain extent, these qualifications are not surpris-ing.
Contrary to design competitions, European tender procedures have never been popular among architects.Almost straight from the start, when the EU direc-tives on public service contracting 2004/ 18/ EG & 2004/ 17/ EG) were introduced, they were associated with bureaucracy and high transaction costs.During the past several years, these frustrations seem to have increased rapidly.In the Netherlands, architects have refused to take part in architect selections, have ended up in court fighting jury rulings and fellow architects, and had the issue addressed in Dutch parliament.
rently, these thresholds are 125.000for services bought by the central government, and for services for other government bodies e. g. provinces, municipali-ties and universities).In practice, this means that EU regulations apply to almost every substantial public design assignment, ranging from schools and town halls to hospitals and prisons, which are all critical compo-nents of the built environment we live in.EU procedures have never been popular among archi-tects and to a lesser extent clients.From the inter-views conducted during this research project, it is clear that they are often considered as bureaucratic, formal-istic and inflexible directives from far-away Brussels.That image, however, is not entirely justified.Close reading of the EU directives shows that they are actu-ally quite open, providing public clients with multiple possibilities for organizing the selection process.The rules even allow the « classic » architectural competi-tion, as long as it is based on anonymous submissions, strict preset evaluation criteria and expert judgement meaning that at least a third of the jury members should be experts on architecture).Thus, it appears that EU regulations actually do not provide a barrier to the creation of a vibrant competition culture.
Interestingly enough, however, Dutch public clients have not shown much enthusiasm for organizing architectural competitions in recent years.Instead of competitions, the majority of the Dutch public sector clients prefer to use the so-called restricted tender procedure to select architects.Statistics show that in the past four years, 89% of all design-related tenders by public clients were EU restricted tender procedures see Tab. 1, Geertse et al. 2009).Open EU procedures, such as a design competition, represent only 5% of the total of design tenders, of which many are ideas com-petitions without any follow up.
The Dutch interpretation of the restricted tender procedure can best be compared to the invited com-petition structure, in which a client invites a limited number of architects to develop and present a design proposal.The procedure consists of three phases: a selection phase, a tender phase and an award phase see Fig. 1).In the selection phase, interested architec-tural firms can submit a request to be invited for the tender.Beforehand, the contracting authorities have to indicate which criteria they will use to evaluate the suitability of a firm and the maximum number of candidates they intend to invite five is the minimum number according to EU directives).In the Nether-lands, this « pre-selection » is most often done by set-ting minimum requirements concerning a firms expe-rience, competences and financial situation.
In the second phase « tender phase » the preselected firms are then invited to supply the client with more detailed information about their project team and their ideas concerning the project, including a pricing proposal for their services.In most cases, Dutch clients will ask all five firms to make a concept or schematic design; sometimes the firms are asked to develop detailed design proposals including scale models and cost calculations.
During the third and final phase « award phase » the client selects a winner who is awarded the con-tract.
The contract can be awarded based on what is called the « most economically advantageous tender » MEAT) or the lowest price only.In architecture, almost all tenders are awarded on the MEAT princi-ple, which means that the client evaluates which firm and/ or which design proposal provides the best « value for money » 4 Different phases, different problems In theory, the procedure described above could work well.Relevant EU regulations ask for little more than a transparent process and the use of criteria that are

2%
Open NL procedure Tab.1: Number of architectural design service tenders from 16.7.2005 to 1.11.2009Anzahl der Ausschreibungen von Leistungen im Bereich Architektur zwischen dem 16.7.2005und dem 1.11.2009Nombre dappels doffres de services architecturaux entre le 16.07.2005et le 01.11.2009Quelle : Geertse et al. 2009 clear, pre-defined, proportional, and as objective as possible.Reality, however, proves to be more difficult, at least in the Netherlands.
As mentioned earlier, this way of tendering is cur-rently being heavily criticised in the Netherlands, espe-cially by architects.An overview of the problems that are associated with the restricted tender procedure is given in Tab.In this paper, the focus is on the subjectively-felt main issues of each of the three phases: 1) Pre-) selection phase: here the main issue concerns the suitability criteria that are used to deter-mine which architectural firms are allowed to tender and which are not.The main criticism is that criteria are too strict, excluding young and small firms.
2) Tender phase: here the main issue is the materials that the architects are asked to produce as « proof » of their competences.The main criti-cism is that clients are « over asking » demanding detailed design proposals while the actual design process has not yet begun.
3) Award phase: here the main issue concerns the judgement of the qualities of the firms and their proposals.The main criticism is that judgement methods tend to be pseudo-objective and com-plex, looking at formalities rather than content.
These three issues are discussed below in more detail and take both practice and theory into consideration.
4.1 The pre-) selection phase: criteria for financial security or design talent?
The pre-) selection phase of a restricted EU tender is focused on limiting the « playing field » The goal is to choose the five or more) most promising and most suitable firms with whom the client wishes to intensify relations during the tender.For this « short listing » pro-cess, clients use so-called suitability criteria.These are qualitative or quantitative criteria concerning a firms competences, experience and/ or financial situation.
The main problem with these criteria is that they tend to be very restrictive.With regard to experience, the majority of clients ask for comparable reference projects that need to have been realized in the last three to five years.For example: a client wishing to build a new school allows only firms in the selection process that have designed at least three other com-parable schools within the past five years.By setting such requirements, young and new design firms do not have a chance of being awarded substantial public contracts.Furthermore, this approach pushes big and medium-sized offices into a particular market segment because they are likely to be more successful in ten-ders for building types in which they already have a strong track record, in this case school design Kempe Thill 2008).
The same problem arises when the client sets high requirements for the turnover of a firm.In most ten-ders, clients set a minimum requirement for the aver-age turnover over the past three years.Such a require-ment The use of restrictive suitability criteria can prob-ably be explained by the clients wish to select expe-rienced and competent firms, fearing that young and small firms will not be able to handle the project, thereby increasing the risk of budget overruns, delays and quality problems.This is an understandable posi-tion because, as in any profession, experience builds competence.But it is also a position that can be ques-tioned because in architecture, experience and finan-cial success are not the same as the ability to design or the expertise to deal with specific complexities in design.Architects are proud of their capabilities to analyse complexities on an abstract level in almost any type building project and to find solutions beyond the obvious.These intellectual competences can be built by experience, but cannot be guaranteed by it.Some architects even suggest the opposite, pointing at projects like the Centre Pompidou and the Sidney Opera House, which were designed by then-virtually unknown architects although one can also argue these are exactly the type of costly projects that have fuelled clients fears of eccentric architects.
To deal with this issue, it is suggested here that clients take a projectby-project approach, making a distinc-tion between innovative projects that require « naïve » entrepreneurship and fresh creativity, conventional projects that need a degree of certainty, and complex projects that ask for the involvement of specific exper-tise and experience.Characterizing projects, however, may be easier said than done..2 The tender phase: looking for a plan or a partner?
In the tender phase, the preselected firms have to deliver more detailed information about their ideas for the project and their project team for the client to reach a final decision.
In the Netherlands there is a tendency towards asking for detailed design solutions in the tender phase, sometimes including cost calculations, scale models and sustainability studies.Architects complain that this way of tendering takes a lot of time and energy although the same can be argued for the classic type of architectural competition.The Netherlands most famous architect Rem Koolhaas commented on this issue Bennett 2007: 1): « We are letting ourselves be drained of endless resources and huge amounts of ideas all to no avail [...].I am on a campaign now to convince the world that this kind of com-petition is hopeless and that we should find more efficient forms of architectural competitions against each other [...].It is amazing how the best brains in the profession can be manoeuvred into a position of complete weakness when they should be in a position of strength » Furthermore, there is a financial reason for the com-plaints about demanding clients.In Dutch tenders, the Negotiations after announcement of the winner financial compensation for architects work tends to be limited, ranging from 2.000 or nothing for small projects like elementary schools, to 15.000 for com-plex projects like a large university building.These amounts do not cover the hours and expertise that these firms put into their design proposals.Obviously, a certain amount of costs is justifiable because it con-cerns an acquisition process, but the input should be in proportion to what firms can bear and in line with the size of the project.The Royal Institute of Dutch Architects estimates that participating in the award phase of a tender procedure costs about 45.000; this is about 10% of the average annual net turnover of an architectural firm ( 543.000 in 2009).For large firms, however, with an estimated average net turnover of million, this is only 0.01% on their net turnover.
These numbers demonstrate that tenders are presum-ably more feasible for large firms than for the average architectural firm.
The root cause of the problems in the award phase may be a lack of clarity about the actual aim of the architect selection: is it the selection of a product a design), the selection of a service activity of a design firm), or the selection of a partner a designer)?
Fig. 2 presents the tensions that often appear from the different interests at play in architect selections.The left side of the figure shows the focus on the product, which is based on the clients intention to acquire a design product.This is what is currently happening in most Dutch tenders and also applies to the classic architectural competition.This approach is based on the assumption that a partner in architectural design can be chosen by judging his or her physical work or a preview of this product.
The right side of the figure shows the procurement principles and process approach.In this approach, architects are considered as entrepreneurial service providers competing for contract.Underlying the EU regulations is the idea of selecting a partner for a building project who is capable of designing the future building.Such a process focuses on maximum value for the client and therefore the client has the final deci-sion authority « he who pays the piper calls the tune » Tensions originating from the competing conceptions of the architect selection process Spannungen, die sich aus der Konkurrenz unterschiedlicher Vorgehensweisen zum Auswahlverfahren von Archi-tekten ergeben Divergences relatives aux conceptions architecturales apparues au cours du processus de sélection Graphics by the authors, based on Volker 2010 In order to know with whom the client will be doing business with, interaction is an important element in the selection process.
Both approaches in Fig. 2 are relevant to selection procedures in architecture.The problem in current practice seems to be that often clients are not aware of their aims.For them the process of architect selec-tion is a process of sensemaking, in which they become aware of their needs and ambitions Kreiner 2006; Volker 2010).The procurement regulations, however, presume that a tender is based on well-articulated ambitions and a detailed brief.The solution to this problem may be to look for a compromise.Instead of asking for detailed design proposals, clients could ask for schematic design proposals in combination with a vision on the process i. e. looking at issues like user involvement, quality management and cost control).
Furthermore, more possibilities for dialogue between the client and the architects during this phase may be of value.Currently, communication tends to be highly formal, consisting of an exchange of written questions and answers.Adding more possibilities for face-toface communication may give the architect a better « feel » for the clients wishes, while clients get a better view of the « soft » qualities of the architect e. g. their ability to listen and their competences in explaining their design ideas).This, however, is likely to make the selective process more subjective.Factors such as the « likeability » of an architect will become more impor-tant while they may obscure the clients view of the architects design competences.
4.3 Award phase: judging based on feelings or figures?
In the award phase, the commissioning body has to review and evaluate all the material that has been submitted by the architectural firms.For many archi-tects this process seems like a mysterious black box of which the outcomes are hard to predict or to influence.
For clients, it tends to be a difficult process in which they have to balance intuitive judgements as well as rational analytical thinking to comply with procure-ment regulations see for example Hogarth 2005; Sin-clair & Ashkanasy 2005).
One of the typical aspects of the current evaluation systems that are used in Dutch selection processes is the rating of different qualities with « points » Each design proposal is assessed against a rather long list of criteria, which are then weighed, and added up to a single score.This system suggests that architectural quality is measurable and scalable, which matches the requirement from EU regulations that evaluations should be as objective as possible.
Perhaps not surprisingly, practice shows that measur-ing quality is difficult.In the study cases underlying this article, only four basic levels of design quality were found: under-performance, basic performance, added value, and design excellence Volker 2010).
Furthermore, it was found that different qualities are overlapping and interrelated.This is in contrast with the objectivity principle that is sought in EU tender regulations, which assumes that criteria are mutually exclusive with clear boundaries and that they together fully cover the award decision.
This assumption is also in conflict with the phenom-enon of compensating judgements in decision making.The concept of the « wow » -factor Desmet et al. 2007), the « surprises » in the model of Kano 1984) and the « thrill » factor of Kreiner 2007) suggests that an affective response to a design proposal can compen-sate deficiencies and ease choice.For the validity of the value judgement, an assessment system for design quality should therefore also incorporate possible compensation, increasing insights, and possible over-lapping qualities.
Obviously, it is not easy to incorporate such complex mechanism in a procedure that is expected to be objec-tive and transparent.One recommendation can be to use more « simple » evaluation systems, with only a lim-ited number of criteria allowing for more integrated judgement) and a limited number of rating levels, accompanied by written explanations avoiding the problem of « measuring » quality).Furthermore, it can be recommended to create a bigger role for domain specific experts.Experts are usually better in complex and integral assessments than non experts Hogarth 2005).It should be avoided, however, that the client gets the impression that they could be overruled by experts.Instead of being the prime evaluator, it is possible to give the expert the role of facilitator, thus excluding them from the judging process, but ensuring their support in helping clients and evaluation com-mittees to reach a clear and sound judgment.
5 Promising developments Fortunately, many of the ideas mentioned above are addressed in a new tender model for architect selec-tions which was developed by an independent nonprofit agency called Architectuur Lokaal.The model, called the « Kompas Light » Architectuur Lokaal 2009b), was launched in December 2009 and it is being endorsed by major client bodies such as the Dutch central government.The digital model guides commis-sioning parties step by step through the relevant forms, resulting in ready-to-use guidelines for invitations to tender.Recommendations help, or urge, clients to put a focus on architectural quality and avoid the use of overly complex and restrictive criteria.
It is important to note, however, that Kompas Light is very instrumental and offers no guarantee for solving the problem of disproportional or excessive selection criteria and vague evaluation methods.For these prob-lems to be solved, the use of this new model has to go hand in hand with a cultural change among clients, or even the industry as a whole, that moves away from the current risk-averse attitude.

Conclusion
This article has shown that nowadays very few Dutch clients choose to organise a classic architectural com-petition.Instead, they prefer to use a so-called restric-tive tender procedure which allows clients to invite a limited number of architects to develop and present a design proposal.
In theory, this process can be quite effective and more efficient than an architectural competition because it limits the number of participants involved.The Dutch interpretation of the procedure, however, could be improved.The most notable problems discussed here were 1) the restrictive suitability criteria in the pre-) selection phase, 2) the extensive design proposals that are required in the tender phase, and 3) the complex-ity of the evaluation methods that are used for award-ing the commission.
In response to these problems, this paper suggests a projectby-project approach.The right type of pro-cedure depends on the type of project, making a dis-tinction between innovative projects that may require « naïve » and fresh creativity, and conventional projects that may ask for certainty and risk avoidance.Selecting an architect for a high profile museum asks for a differ-ent approach than selecting a designer for a mediumsized school with a tight budget.Furthermore, it is suggested that procedures should be made less formal, providing more room for naturalistic and integrated decision making, thereby closing the gap between the formal evaluation systems, as described in the official tender documents, and the messiness of reality.
Another general recommendation concerns the involvement of experts.Currently, many selection pro-cesses are led by purchasing departments or tender consultants who may be highly knowledgeable on tender procedures but cannot support inexperienced clients evaluating the qualities of architectural firms and design proposals.
It is hoped that these recommendations, and new tools like Kompas Light, can help to re-establish the healthy architectural climate for which the Netherlands was once famous.enquiries of the creative industry).Den Haag: Senter Novem.
Summary: Dutch design competitions: lost in EU directives?Procurement issues of architect selections in the Netherlands This paper discusses architect selections for public commissions in the Netherlands.It shows that Dutch public clients seldom use « classic » design competi-tions to select architects.Instead, they use so-called restricted EU tenders which are associated with high transaction costs, disproportional suitability criteria and time-consuming procedures.These tenders are thought to lead to the exclusion of young architects and further specialisation in the architectural profes-sion, which may, eventually, result in predictable or even mediocre architecture.
Based on a literature review, personal experience and case studies, the paper identifies three main problem areas: 1) the use of overly strict suitability criteria in the pre-selection phase; 2) the extensive design pro-posals that are required in the actual tender phase, and 3) the complexity of the evaluation methods that are being used for awarding the commission.
After analysing these problems, the paper offers suggestions for improvement and briefly presents a promising new model, arguing that a more open, qualitative and process-oriented approach is likely to create tender procedures that are more efficient and effective.