Selectionism vs Trial and Error in Large-Scale Development Projects
by Dr. Mohammad Baydoun, Visiting Research Fellow, Project Manager Millennium Development International and Lecturer American University of Beirut
During the last two decades, the development sector was booming worldwide, especially within developing countries that are rich in natural resources. This boom increased pressure on governments to develop large-scale urban projects that can contain the newly emerging developments (Sagalyn 2007). The purpose of these projects is to divide large land areas into smaller parcels that meet demand requirements while providing necessary infrastructure for these parcels. In order to secure success of urban development projects, it is essential to mitigate or reduce the impact of its risks, especially that its processes usually entail a large number of unforeseen risks (Kyvelou and Karaiskou 2006).
The intellectual discourse between selectionism and trial and error theory has been widely discussed in the context of technological projects that are complex and entail a lot of unforeseeable uncertainty. In this article, we shall bring this discourse to the context of large-scale development projects that are known to be highly complex and risky in order to understand how these theories could inform the risk management approach to be followed for these projects. The Shamiyah project in Saudi Arabia was selected in order to analyze the nature of risk and the applicability of both selectionism and trial and error theories within this context.
The research relies on a case study enabling to qualitatively investigate the practiced approach for risk management of this particular project and to illustrate the developed theory using examples from a significant and representative case (Yin 2009). The case study, which is the Shamiyah project, was selected based on its largeness of scale, international recognition and coverage of a wide spectrum of development phases. The Shamiyah project aims at the redevelopment of a 1.38 square kilometers hilly site - the Shamiyah District - facing the northern edge of the Haram Al-Shareef (the great mosque) in Makkah. The project is composed of three zones.
Tushman and O'Reilly 1997 and McGrath 2001 claim that considering different variations is the ultimate way for achieving novelty. Selectionism may include multiple deviations for the same solution that are gradually narrowed down to the most convenient solution (Ward et al. 1995; Sobek et al. 1999), or multiple solutions that are tested before selecting the ultimate one (Stalk and Webber 1993).
Flipping to trial and error learning, Chew et al. 1991 and Lynn et al. 1996 promotes trial and error as a mean for continuous learning and highlight the importance of flexibility in responding to unprecedented situations through knowledge making. In addition, project risk management promotes continuous learning and adaptation (Chapman and Ward 1997; Miller and Lessard 2000). The essential issue for the success of this approach is to adjust the project plan based on new data received during the course of project, whether this data is actively gathered or passively received.
The studies of the Shamiyah project were found to be arranged along two major tracks: technical and financial. The aim of the technical track was to come up with a project that is valid from different technical angles. For this purpose, the technical studies were managed simultaneously to make sure that each study meets the highest professional standards, and that all these studies are synchronized and folds within a single entity defining the technical dimension of the project. Hence, simultaneous management of all these studies that varies in required expertise was highly challenging and involved a high level of risk. In addition, a balance needs to be achieved between requirements of different parties. The requirements of this balance are not a 'blue print' solution that can be adopted in different projects. Hence, solutions are to be proposed on case by case basis.
The technical track extended over four stages (Figure 1). The first stage entailed preparing six master plan options for the sake of exploring different options that could provide the ultimate development strategy for the project. These schemes were developed based on the challenges and risks that were know in that moment of time and included in the design brief shared with the consultants. After assessing the six schemes, it was found that three options hold value and it was decided to further develop these three options before deciding on the ultimate solution. Based on the assessment of the six schemes and on data received during the course of development of the six schemes, the projects challenges and risks were revised as new information on the project was available. The revision of the challenges and risks were the basis for development and revision of the three schemes. After conclusion and assessment of the outcome of the second stage, it was decided that two competitors shall join forces to develop a single master plan as each of the schemes prepared by those two consultants held its own merit. In the stage to follow, one consultant was selected to develop the scheme. In the latter two stages, the scheme was continuously revisited based on information and feedback received from different stakeholders. The followed approach was perceived to be allowing for identification of the ultimate solution for this highly complex project that holds a large number of uncertainties. Due to its large-scale and number of involved disciplines and systems, the project was highly complex. In addition, as the project was unprecedented in its context in terms of scale, a high number of unforeseeable risks were found.
Figure 1: Process followed for studies of the Shamiyah project.
The financial track was mainly related to conducting a market study to validate the project program and identify different cost and revenue assumptions to be used in the financial model. This track was run in parallel with the technical one. The formulation process of the project program focused on maximizing capitalization on allowable built-up area while running in parallel the market study that was used to test the business logic of the project. This is due to the fact that demand in Makkah on accommodations in general and hotels is particular is extremely high because of the increase in number of visitors each year and the obvious lack of competition as Makkah is the only pilgrimage destination for the Muslims. This high demand encouraged the developer to initiate design studies to investigate possible options without waiting for results of the market study. The scope of the market study was basically to identify high level economic indicators and cost and revenue assumptions. The high level economic indicators helped in analyzing profile of users in order to accommodate for their needs in the master plan. Despite the fact that the uses were set based on regulations, the typologies of units for each use were defined based on the outcome of the market study. In addition, the cost and revenue assumptions were investigated to provide necessary assumptions for the feasibility study that helped in optimizing the return rates of the project and assessing the feasibility of the developed design schemes. Furthermore, the feasibility study helped in confirming that the formulated project program based on the maximum allowable built-up area was financially feasible. Based on the outcome of the technical assessment for design solutions and the preliminary feasibility study, three alternatives were developed in the second stage, during which a second run for the feasibility study took place. During the final two stages, during which the final design solution was formulated, two additional runs for the feasibility were conducted in order to continuously monitor the financial performance of the scheme and to identify the expected return on investment for the final design. This track also entailed a number of unforeseeable risks.
Applicability of selectionism and trial and error theories in large-scale projects
Risks that were faced in the case study are mostly unforeseeable due to its complexity and uniqueness, which renders databases generated out of previous projects obsolete. This was found to be similar to the context of identified literature, in which selectionism and trial and error theories were discussed. Nevertheless; the analysis of the nature of the identified risks lead to the conclusion that unpredictability in the context of the case study is twofold: unforeseeable in kind and unforeseeable in impact (Figure 2).
Figure 2: Conceptual applicability of the selectionism and trial and error in the context of large-scale development projects.
The first category is related to risks that were not predicted from the beginning of the project as there were no early indications for the presence of these risks. Risks like unreasonable regulations and unprecedented increase in construction cost that were explained in the earlier section could not be predicted. The first risk could not be identified till the design is already underway and the second was an unusual event that could not be traced based on recent trends.
The latter category is related to risks that could be foreseen in terms of likeliness of occurrence while the impact of the risk when realized could not be identified. Risks like lack of design for crucial infrastructure components, qualitative regulations and large-scale of the project were considered as potential risks without identifying the impact of these risks. As an example, the presence of qualitative regulations was considered as alarming from the early beginning; nevertheless, the impact of this risk could not be predicted early on as the project manager cannot know how the public authorities will use such regulations. In addition, the risk of land prices was known from the beginning; yet, the impact of such risk on the project could not be known prior to completion of properties evaluation as Makkah is a unique place and the adopted method of evaluation was also unique.
These two different categories of unpredictability combined with the complexity component of large-scale development projects have implications on the approach to be selected for risk management. In first category, the full blindness of the risk implies adopting the trial and error approach as in case of adopting parallel tracks, the number of such tracks will be infinite as parameters are not available to define the most likely scenarios in such a case. For the latter category, selectionism would make more sense provided that no budget constrain is present as this will allow the managers of these projects to test different possible solution based on predefined scenarios in order to select the most convenient solution. As both risk types were found within the same project, the structure within which these two approaches could be applied was formulated based on analysis of findings.
As per the findings, the studies went through multiple stages that could be conceptualized as follows: multiple options, narrowed options and single option stages, see figure 3. The scenarios of the first stage were formulated based on risks and challenges defined at this stage. Identified risk were known to be a possibility without identifying the impact of these risks as large-scale projects hold a high level of uniqueness that renders databases generated out of previous of projects obsolete. In addition, the selected scenarios were revised before launching the narrowed options stage as more data became available, which allowed for enhancing the scenarios and even drastically changing some of them. Accordingly, despite the obvious application of selectionism during the first two stages, the definition of the scenarios involved the trial and error approach as the scenarios were revised based on newly available information by the end of early stages. During the last stage, a single scenario is adopted; yet, this scenario was continuously revised based on received information from different stakeholders and changes in the project context whether political or economic. Therefore, the adopted practices for this stage fall under the trial and error approach.
Figure 3: Structure of application for selectionism and trial and error in the context of large-scale development projects.
In brief, both approaches that are discussed in literature for risk management of complex projects with high level of uncertainty are applicable to the context of large-scale development projects. The applicability is dependent on the nature of the risk whether it is unforeseeable in kind or in impact while satisfying the complexity requirement in both cases. In addition, even in the multiple-option stages, in which the selctionism approach is adopted, the definition of the scenarios involves the trial and error approach. The trial and error approach is also adopted in the final stage of the design involving the development of a single solution derived out of lessons learned in the previous stages.
This article contributed to the body of literature related risk management of large-scale development projects. It brought the intellectual discourse related to selectionism and trial and error theories to the context of large-scale development projects and added a differentiation aspect on this front by distinguishing between risks that are unforeseeable in kind as opposed to those unforeseeable in impact. This article shall be considered as a trigger for further development of risk management theories within the context of large-scale development projects, more so since literature still contain a number of gaps on this particular topic, especially in terms of models that could make current practices related to selectionism and trial and error more efficient.
The article is a summary of the following published paper: Baydoun, M. (2013), 'Applicability of Selectionism and Trial and Error in Large-Scale Development Projects', PM World Journal, 2(7): 1-17.
Chapman, C. B., S. C.Ward. (1997), Project Risk Management: Processes, Techniques, and Insights, John Wiley & Sons, Chichester, U.K.
Chew, W. B., D. Leonard-Barton, R. E. Bohn. (1991), 'Beating Murphy's law', Sloan Management Review 32:3, pp. 5-16.
Kyvelou, S. and Karaiskou, E., (2006), 'Urban development through PPPs in the Euro-Mediterranean region', Management of Environmental Quality: An International Journal 17: 5, pp. 599-610.
Lynn, G. S., J. G. Morone, A. S Paulson. (1996), 'Marketing and discontinuous innovation: The probe and learn process', California Management Review 38:3, pp. 8-37.
McGrath, R. G. (2001), 'Exploratory learning, innovative capacity, and managerial oversight', Academic Management Journal 44:1, pp.118-131.
Miller, R., D. R. Lessard. (2000), The Strategic Management of Large Engineering Projects, The MIT Press, Cambridge, MA.
Sagalyn, L. (2007), 'Public/private development', Journal of the American Planning Association 73: 1, pp. 7-22
Sobek, D. K., II, A. C. Ward, J. K. Liker. (1999), 'Toyota's principles of set-based concurrent engineering', Sloan Management Review 40, pp.67-83.
Stalk, G. Jr., A. M. Webber. (1993), 'Japan's dark side of time', Harvard Business Review 71:4, pp. 93-102.
Tushman, M. L., C. A. O'Reilly. (1997), Winning through Innovation: A Practical Guide to Leading Organizational Change and Renewal, Harvard Business School Press, Boston, MA.
Ward, A., J. K. Liker, J. J. Cristiano, D. K. Sobek II. (1995), 'The second Toyota paradox: How delaying decisions can make better cars faster', Sloan Management Review 36:3, pp.43-61.
Yin, R. K. (2009), Case Study Research: Design and Methods, 4th edn, Thousand Oaks, CA: SAGE Publications Ltd.