Identifying the Factors Driving the Uncertainty in Transport Infrastructure Project by Application of Structural Dynamic Analysis to a Backcast Scenario Pete Sykes , Margaret Bell, Dilum Dissanayake Newcastle University, School of Engineering, Cassie Building, Claremont Road, Newcastle, NE1 7RU, UK Abstract Transport planning, in theory, is underpinned by rational analysis of the benefits of proposed developments. However, project outcomes do not always follow the results of that analysis and uncertainty is evident during the decision making processes. This research has devised and demonstrated a method to analyse that uncertainty, focussing on the early stages of the project lifecycle. Stakeholders were interviewed to elicit their opinions about a normative scenario and these interviews coded using qualitative data analysis techniques. The emerging variables were analysed, using a structural dynamic model, based in complexity theory, which develops measures of connectivity to classify variables by their roles in inception and uncertainty in the project. The case study was based on a disused railway with contradictory views on the benefits of reopening it. In the normative scenario, the rail service is re-instated in conjunction with a new sustainable urban development. The findings from this case study were that executive leadership and collaboration between Local Authorities were the most influential determinants for progress, and that the prime causes of uncertainty were the extant economic and planning policies. During the course of the project, structural governance developments have occurred in the UK that have endorsed these findings. 1. Introduction Transport planners design transport systems, be they road, rail, air or sea. On the face of it, the task should be straightforward: Identify a transport need, design and assess a solution to meet the need to move people and goods efficiently, bid for and win the funding to build it, then, once built, to use and manage it. There are, however, many examples of transport projects which have a long gestation period before coming to fruition, and some never materialise at all. High profile developments such as a third runway at Heathrow airport (Grekos, 2014), HS21 (Divall, 2017) and the Aberdeen Western bypass (Transport Scotland, 2017) suffer lengthy procrastination lasting decades with deferral and delay seeming to be expected at every stage of the approval process. There are well documented processes for assessing transport developments (DfT, 2016; Worsley and Mackie, 2015; Ortuzar and Willumsen, 2011; Berechman, 2009; Nijkamp et al., 1998). A transport model is calibrated to represent the traffic of today; adjustments made to reflect planned changes in travel demand, to the network itself or to the control systems used to manage movement within it; and the effect of those changes is quantified by the model outputs. The economic benefits are subsequently assessed (Banister and Berechman, 2000) and used by local authorities, national 1 High Speed Rail Line 2 1 authorities and politicians to guide their decision making as they allocate funds from local and national budgets to develop the transport network. However, a study by Welde et al. (2013) revealed that the correlation between projects selected for funding and their estimated value, can be weak and indicates that the result of the cost benefit assessment exercise was not necessarily the key factor in project selection. Wachs (1985) and Vigar (2017) discuss the significance of the political dimension to transport decision making. Wachs (1985) comments that the majority of the research has been in the technical aspect of transport assessment and that more is needed on the social and political dimensions and Vigar (2017) argues that while successful project implementation requires technical, local, and empirical knowledge to support the decision, political acumen also is required for the decision to proceed. The transport decision process, therefore, consists of much more than the rational analysis of an individual transport problem and developing a suitable solution. Transport project uncertainty conforms to what Lindblom (1979; 1959) describes as “Muddling Through” as transport developments find their niche in a large complex ecosystem of collaborating and competing policies, multiple infrastructure developments, and changing patterns of transport use. It is what Rittel and Webber (1973) describe as a “Wicked Problem” one which defies rational analysis and evolves as it is analysed. The goal for the research presented here, therefore, was to devise and trial a novel analytical method to provide an insight into the causes of uncertainty in a proposed transport project and to provide this insight in the entire scope of the proposed project without limiting the analysis to the quantitative assessment stage. While there have been studies into the structural and stochastic uncertainties inherent in the transport model itself, there have been few attempts to look at the whole process from concept to fruition. The research gap is succinctly identified by Marsden and Reardon (2017) “We need to not only be able to map the decision making systems and formal structures of power but also recognise the more informal networks and sub-systems of actors that coalesce around policy issues…. there is a need to engage with substantive questions of governance which pay greater attention to context, politics, power, resources and legitimacy”. Therefore the goal, in this research, is to gain an understanding of the uncertainty within the decision making process of a transport project. Also, explicitly, there was no requirement to contribute to the decision as to whether or not to proceed with the proposed development, hence freeing the research to investigate other forms of modelling not commonly associated with transport assessment. A detailed description of the methodology, the derivation of the parameters used in the analysis and a study of the sensitivity of the results to those parameters is described in detail in Sykes et al. (2018). This paper focusses on the scenario analysis, and the findings from the case study. 2. Review of Uncertainty and Complexity This review is divided into two sections: The first section reviews the uncertainty in the assessment processes with an emphasis on scenario planning to assist in managing uncertainty. The second focusses on complexity theory and techniques used to identify uncertainty. 2.1 Uncertainty in the Assessment and Decision Processes Before discussing approaches to managing uncertainty, we must first discuss the nature of uncertainty itself. Investigations into the sources of transport planning uncertainty implicitly refine 2 the concept according to the class of uncertainty under discussion (Berechman, 2009; de Jong et al., 2007; van Geenhuizen and Thissen, 2007; Kikuchi, 2005; Refsgaard et al., 2005; Walker et al., 2003; Courtney, 2001; Rowe, 2001; Funtowicz and Ravetz, 1990; Morgan and Henrion, 1990). In creating a taxonomy of uncertainty, researchers identify three major areas of modelling and assessment. The first is that which can be dealt with analytically such as stochastic variance or parameter sensitivity and is found in the attributes of the model; namely its algorithms, parameters and data outputs (Saltelli et al., 2008 ; Cacuci et al., 2005; Morris, 1991). The second is the incompleteness of the model: Walker (2003) describes this as the uncertainty due to the incomplete representation of behaviours and relationships in the model. Mattot (2009) refers to this as model technical uncertainty, stemming from erroneous knowledge or incomplete models. Rasouli and Timmermans (2012) see it as oversimplification or incompleteness of the model, and both Morgan and Henrion (1990) and Rodier (2007) comment on uncertainty as deficiencies in the functional form of a model i.e. in trip choice and in driver behaviour. The third category of uncertainty, identified by researchers, is in describing the environment for the proposed development. There is an extensive body of literature and text books describing the techniques used to develop and deploy scenarios to form a framework to assess future options (Chakraborty, 2011; Giaoutzi et al., 2011; Godet et al., 2009; Lindgren and Bandhold, 2009; Marchais-Roubelat and Roubelat, 2008; Wright et al., 2008; Harries, 2003; Peterson et al., 2003; Chermack et al., 2001; Godet, 2000; van der Heijden, 1996; Porter, 1980). However, the word scenario, is overloaded in the transport planning literature and is variously used to refer to a range of forecasting techniques: from a selected list of predetermined options with uncertainty limited to that which can be described by the possible ranges of a few quantifiable variables (i.e. the future fuel price or a range of growth forecasts), to a descriptive sample of plausible futures extrapolated from the present in a study designed to cope with a wider scope of uncertainty (i.e. the impacts of emergent technology based intelligent mobility solutions). Backcast scenarios are described by Dreborg (1996) and Robinson (2003) as a scenario study which goes beyond what is possible when forecasting from the present. Backcasting studies employ explicitly normative scenarios and are concerned with the route (or routes) to reach a stated goal working backwards from that goal to the present. Backcasting is not designed to facilitate discussion on a range of futures, but instead to examine the interplay and relative effect