1 Identifying the Factors Driving the Uncertainty in Transport

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 of the component sub goals and policies that form a normative scenario which describes a desired outcome. Marchau and van der Heijden (2003) describe the same process as rendering an image of a desired future and identifying the path to it. If the path cannot be found, then the image cannot exist and therefore must be adjusted. Examples of the use of backcasting to provide a framework to stimulate creative activity amongst workshop participants can be found in the VIBAT (Visioning and Backcasting for Transport for London) project (Hickman et al., 2009) to evaluate the robustness of a package of carbon reduction policies under different future scenarios and hence refine the choice of options in the policy package, and in Soria-Lara and Banister (2017) to examine the efficacy of different combinations of transport related policy packages and their political feasibility.

2.2 Complexity and the Structural Dynamics Model Weaver (1948) is widely referenced as the source of research in complexity theory by introducing the concept of “organised complexity” to problem definition. Subsequent work in complexity gives a multiple set of definitions of a complex system as one which is (1) self-organising with emergent behaviour; (2) adaptive as the system evolves; (3) crosses functional disciplinary boundaries; and (4)

3 is unpredictable as small changes may have large effects (Colander and Kupers, 2014; Holland, 2014; Cairney, 2012). Flood and Carson (1993) discuss complexity in terms of systems theory, given a concrete foundation in systems dynamics models and Byrne and Callaghan (2014) describe applications of complexity in social sciences. The emphasis in Flood and Carson (1993) is on analysis of the connectivity of interacting elements identified in the system and the boundaries of the system. Cairney (2012) reinforces this view on understanding connectivity as a precursor to understanding complexity “ as a network of elements that interact and combine to produce systemic behaviour that cannot be broken down merely into the actions of its constituent parts”. Cairney (2012) regards complexity theory as having value in understanding political choices made on intractable policy problems where rational analysis fails. In such situations, there are many variables with multiple interrelationships many of which are uncertain or ambiguous, an observation which can be seen in the decision making process of a transport planning project. One direct approach to quantifying uncertainty through an analysis of connectivity is embodied in the structural dynamics model which is similar to the causality diagrams embodied in a systems dynamics model. The structural dynamics model was developed by Gordon, Godet et al. and Vester (Vester, 2012; Godet et al., 2009; Gordon, 1968). It fundamentally infers that complexity and uncertainty are synonymous with connectivity and is expressly designed for the class of analysis described here, to identify the drivers of uncertainty in a project. A structural dynamics model is defined as one which uses a cross impact matrix to construct a network of causalities (where X influences Y) and, by studying those causalities, to identify the key elements that control the system’s evolution. The model was initially created by Gordon (1968), subsequently, Vester (2012) developed the methodology with a goal to understand the system under analysis more than the outcome and Godet (2011), in similar work, added the use of indirect links in a causality chain. In examples of the use of the method, Cole (2006) analysed the issues in developing policy for a water catchment area, Amaya Moreno et al. (2014) studied the interaction between genes to identify the most active, reactive, buffering and critical genes in the network where the precise interactions are unknown and Muric et al. (2013) used a similar method of evaluating connectivity to assess criticality in IT networking hardware. The output of a structural dynamics model is a two dimensional graph of influence and dependency where Godet (2011) describes the elements that fall into the upper right quadrant being both highly influential and highly dependent and are the likely cause of instability in the system. Vester (2012) further categorises elements of the systems under analysis by their position in this space as shown in Figure 1, where those in zone 1 are the most influential and crucial to initiating the system, those in zone 2 are highly influential and also highly dependent and therefore these are held to be the critical variables and the drivers of uncertainty in the system. Zones 3 and 4 contain the indicators of system outputs while zones 5 and 7 contain the sluggish indicators, and the weak control levers respectively. Zone 6 holds those elements which are least important. Finally, the neutral zone in the middle contains the controls which regulate the system. Hence by examining the location of elements in this space, roles can be assigned to them. A structural dynamics model is, at a high level, similar to a systems dynamics model, based on causal loops (Sterman, 2000; Checkland, 1999), however, in a structural dynamics model, entities are not assumed to be quantifiable and links between entities are not assigned a positive or negative sign (where an increase in A causes an increase or decrease in B). A structural dynamics model may be formed from similar causal chain links as are found in systems dynamics model, but the structural

4 dynamics model offers more flexibility in the qualitative, not necessarily quantitative, nature of the entities it encompasses albeit with a more restricted purpose; solely to analyse the roles of the entities that form the model. 3. Designing the Methodology

The method analyses the uncertainty in a project from the inception stage, when the proposed transport development must find its niche in the current and future policy and infrastructure ecosystem, to the approval stage given that decisions are be made in a complex and highly interconnected environment. This concurs with observations made by Wachs (1985) and by Marsden and Reardon (2017) that research into the social and political dimensions of transport planning is desired and extends the research requirement into analysing what Walker et al. (2003) refer to as scenario uncertainty, and van Geenhuizen and Thissen (2007) refer to as system boundaries and outcomes. The strategy was to adopt a generic modelling approach. First, the analytical requirements were specified. Second, a means of describing the system with a model, including the data, algorithms and required outputs was devised. Third, given the data requirements were now known, the sources and methods to gather, code, and validate that data were detailed. Finally, the analysis was undertaken and the results interpreted.

3.1 Analytical Requirements The outputs required of the analysis are a list of the components of a planning project that most contribute to the progress and to the uncertainty in the project. This must be accompanied an understanding of how they reach their positions on this list. Therefore, the primary input must be the full list of variables governing the assessment and decision process. The task of generating that list must also be a part of the data gathering exercise as an analyst investigating the uncertainty in the project has to avoid imposing the constraints that would occur if what should, or should not, be on that list was pre-determined. This open ended data gathering requirement addresses one of the components of uncertainty; that of model completeness.

3.2 Model The structural dynamics model was selected as the tool to be used in uncertainty analysis. However, one weakness in the documented examples of its use is the reliance on focus groups or a single workshop to generate both the lists of variables and the influences between them. This leads to issues of groupthink, or of the potential dominance of one of the team members - as identified by Booker and McNamara (2004). Indeed, recognising the different views of multiple stakeholders is held to be one aspect of identifying uncertainty (Scolobig and Lilliestam, 2016). In applying the structural dynamics model, Gordon (1968) and Cole (2006), note the problem but offer no solution and Godet (2011) too assumes a single view homogenous view of the system across all stakeholders. Therefore, in this research, the model was extended to account for the causality perceived by individual stakeholders holding their data independently and combining it as required. The model was also extended to use indirect links, as described by Godet (2011), but with a diminishing weight applied as the level of indirection increased. To evaluate the Influence and Dependency for each variable in the structural dynamics model, the sum of the link strengths in the chain of causality is summed. For example, Figure 2 shows a sample

5 causality matrix as derived from the scenario based elicitation interviews. In this example, one stakeholder states that variable-1 influences variable-3 and 3 stakeholders state that variable-2 influences variable-1, etc. First a strength for each link is evaluated using the formula S= Np where N is the number of stakeholders making the link and p is a weighting parameter. Next, this link strength is then weighted using the function w(d) = e-λd where d is the depth of this link into the causality chain between two variables and the λ parameter controls how the link strength diminishes with increasing depth. Figure 2 also shows the same table represented as tree of influences, three levels deep where the width of the line indicates the strength of causality (Np) and the grey shade, its weighting in the tree. Figure 3 shows influence tree for one variable, selected from Figure 2. Here, the Influence value between any two variables is the sum of the weighted strengths in the causality chain between them. Similarly, each variable in the chain also has a dependency value calculated, being the sum of the causality chains leading to it. The derivation of the parameter values used in the Leamside Line case study and the sensitivity of the findings to changes in these values can be found in Sykes et al. (2018). In this case study the values selected were p=0.5, λ=5, and the maximum depth of causality=5.

3.3 Data Collection The data collection method used elicitation from multiple stakeholders, based on a backcast scenario exercise. Scenario planning exercises are primarily designed to be participative, to engage stakeholders in a shared learning exercise, but this does require significant and simultaneous participation of stakeholders and the engagement of potentially many different organisations. In an active ongoing project this may be possible but in a nascent project with strong buy-in from some stakeholders, but little or no buy-in from others, it is difficult to see how this would be achieved without significant sample bias. The method designed here therefore, implicitly, could not rely on multi-stakeholder workshops or similar large participatory exercises. The method had to be able to be applied through contact with stakeholders on an individual basis. Indeed, understanding the different views of the multiple stakeholders is held to be one aspect of identifying uncertainty which requires avoidance of the problems of motivational bias and group-think as identified by Booker and McNamara (2004). This placed a further requirement on the implementation of the structural dynamics model for this research: that it be capable of holding the opinions of the multiple stakeholders separately and be able to combine them. In this research therefore, the process was to create a backcast scenario, then use it as the basis of an elicitation exercise based on one–to-one interviews with a diverse set of stakeholders. The recordings and transcripts of the interviews were then coded to extract the variables from the interviews and the causality between them. Coding is the process of identifying segments of text, data, or media as relevant to the subject such as a theme, concept, or physical object. The goal of a coding scheme is to identify the entities (or variables) in the project and to describe them such that they are coded consistently across different data sources and media. Lewins and Silver (2008) describe coding schema as either inductive, where the list of entities develops as interviews are coded, or deductive where the list is pre-determined before coding starts. In this research, the coding scheme was initially deductive for the physical elements of the scenario given that the topics of the interview conversation were prescribed.

However, the development of the coding scheme necessarily became inductive for the resulting conversations regarding causality. The interview technique was intended to be open-ended, there was no prior knowledge of each individual’s opinions nor any desire to constrain their responses to a fixed set of variables. Indeed, the ability to widen that set of variables depending on responses was one of the goals of the coding method. Links between variables were the formed by searching the transcripts for link words such as “because”, and “therefore”. Causality, and the direction of causality cannot be inferred from simple adjacency of variables, a linking phrase or a link word must be present such as “X is true because of Y” or “A is true, therefore B happens”. Similarly, links were identified in the causal diagram snippets using the lines drawn between each variable in the interviewer’s written notes.

3.4 Uncertainty Analysis The final stage of the method is the uncertainty analysis when, using the structural dynamics model, the variables that are the most influential, most dependent and, as a combination of these two measures, contribute most to the uncertainty in the system were identified and the reasons for this status examined in detail by referring back to the interview transcripts and notes. Also, as the variables were labelled by category, examination to identify which category was most influential, dependent, or leading to uncertainty was possible. Furthermore, the relative importance of links can be quantified by their “potency” – being the sum total of causality flowing through a link, and hence links between variables could be broken to observe how the system reacts to changes to those links which give significant variables their positions in the Influence-Dependency space. Ultimately this analysis could enable a project owner, responsible for realising the proposed scenario, to identify ways of reducing uncertainty by, for example; breaking specific links by acting to remove an influence, or to identify and act on the concerns of one or more stakeholders whose contribution to the causalities is shown to be a significant source of uncertainty. In short the method described here offers the opportunity not only to identify the variables that are the drivers of uncertainty, but to identify why they have this status and how it may be altered.

3.5 The Integrated Method Therefore, the novel integrated methodology developed for this research therefore was designed to link back-cast scenario analysis with multiple stakeholders and carry out analysis using a structural dynamic model to determine the drivers of uncertainty. This is shown conceptually in Figure 4, an overview of the four steps in the integrated method where the stages are as follows: Scenario: The problem in the case study was scoped and the normative back-cast scenario was prepared through consultation with stakeholders who shared a common view of the ideal outcome. Elicitation: A further set of stakeholders was identified. Open ended, one-to one interviews were conducted to elicit their views of the route to the outcome described in the scenario. Processing: The noun and verb phrases that describe the key elements and actions identified in the interviews were extracted using an inductive coding scheme that allowed new variables to be added as the coding proceeded. The causality links between the variables were coded in the structural dynamic model and held separately for each stakeholder interviewed. Analysis: The Influence-Dependency graph was then produced. This stage included a review of (a) the selection of the stakeholders to ensure all relevant points of view are covered and (b) the

7 evolving coding scheme. These reviews are illustrated by the feedback arrows in Figure 4. Finally, the uncertainty analysis, based on the Influence-Dependency graph augmented with the notes from stakeholder interviews and examination of the causality links (illustrated by the forward arrows in Figure 4), was undertaken. 4. Case Study

The Leamside line is a disused railway line between Newcastle and Durham some 5 – 10 miles East of what is now the current East Coast Main Line (ECML). It was opened in 1839 and was part of the North East Railway Company trunk route until 1872 when the current ECML was built. It was closed to passenger transport in 1964 and to freight in 1991, when the line was then mothballed. The UK Transport Act 2000 requires each Local Transport Authority (LTA) to write a Local Transport Plan (LTP) to describe its transport strategy for the following five years and to update that document every five years; guidance on the requirements of the LTP documents was issued in 2008 (DfT, 2009). As part of the Self Conserving Urban Environment (SECURE) Project (Bell, 2013) investigating transport based sustainability inefficiencies, the Local Transport Plans of four Local Transport Authorities in the North East of England (Northumberland, Tyne and Wear, County Durham, and Tees Valley) were studied and as a result of that study, augmented by associated consultancy reports, the Leamside Line was identified as a suitable case study for this research project. The Leamside Line is discussed in two of the Regional LTP documents (Tyne and Wear Integrated Travel Authority, 2011; Durham CC, 2010). Durham CC see the Leamside line in two lights; one to add rail resource to mitigate road congestion, and the other as a freight resource to relieve pressure on capacity for mainline passenger services. In Tyne and Wear, the motivation for re-opening is more based in light rail to augment the existing Metro. Five relevant consulting reports on the Leamside Line (AECom, 2010; Network Rail, 2010; AECom, 2007; Network Rail, 2007; AECom, 2006) and a further report which discusses rail projects in the UK (ATOC, 2009) also were consulted. Within these documents, there are wide ranging diverse recommendations for the mode of use and benefits of re-opening the Leamside Line. The modes of use are variously described as local rail, national rail, dedicated freight and freight diversion only; these modes are either compatible or mutually exclusive depending on the report. In general the business case is found to be weak for a single mode of use, but stronger if multiple modes are proposed, or if the further economic benefits of urban agglomeration are taken into account. In summary, the Leamside line can be seen to present a multi-faceted set of uncertainties derived from the varying expectations of the multiple stakeholders. Each one making their own case for re- opening the line, but without a convincing case clearly identified by any one stakeholder. The Leamside Line therefore was selected as a suitable case study.

4.1 The Normative Scenario The scenario was written by interviewing four stakeholders with an interest in, or knowledge of, either the Leamside Line or sustainable transport in the North East of England. All had a bias towards opening the line though all had different approaches to identifying the benefits of seeing it reopen. A selection based sampling scheme was used to identify the stakeholders and the goal was to cover passenger rail use, freight rail use, and sustainable transport development. This was aligned with the scoping statement which was primarily concerned with re-opening a disused railway. The problem

8 bounds were set to consider the re-opening of the Leamside Line, however, no limits were set on the direct or indirect consequences of this. The notes and recordings from the interviews, typically 1 - 2 hour long, were analysed and a three page scenario written with vignettes to illustrate the narrative. This was then peer reviewed within the context of the SECURE (Self Conserving Urban Environment) project (Bell, 2013) at a workshop involving some 30 academics and practitioners. The scenario focussed on the Leamside line in 2030 providing light rail services to the region and being a part of a sustainable transport network linking the major cities having been re-opened in 2025. The Leamside Line, in conjunction with the Tyne and Wear Metro, is also used for local freight movements with road transport for the “last mile “delivery only. In parallel with the re-opening of the Leamside Line, the existing small village of Fencehouses has been expanded into a small town, with a strong emphasis on sustainability in its design and in its links to the local transport network. The scenario presents vignettes describing life in Fencehouses, the use of the renewed Leamside Line in local freight handling and the personalised view of a scheduling manager on the mainline railway describing how the task evolved as the Leamside Line was opened. The scenario is presented in Appendix A.

4.2 Data Collection The first batch of stakeholders were identified from the project’s professional and academic network. This network is composed of Transport Planners, Transport Modellers, and Academics. The preliminary results were presented at the final conference and review meeting of the SECURE project (Bell, 2013) and following discussion in the conference forum, more stakeholders were identified in areas of transport user management, sustainability activist and intelligent transport systems. Table 1 summarises the characteristics of the stakeholders noting that: there is good coverage of Local Authority staff and consultants, with a prevalence of planners or planning advisors. Also, all stakeholders either have specific knowledge of the Leamside Line, or relevant knowledge of the services it may provide. In total, eleven stakeholders were interviewed between 2013 and 2014. The interviews were recorded and notes were taken as the conversation continued. The interviews were later coded to extract the relevant variables which were classified into a small number of categories: Economy and Demographics, Fencehouses, Leamside, Politics, Public Transport, Road Travel, Sustainability, Freight Links between variables were formed by searching the transcripts for link words and phrases between adjacent variables. A typical example of the text that forms links between variables is as follows: “disconnected highway network” because “… historic Co. Durham villages” and “Not located on good nodes and links” therefore “they use private cars”. These links were interpreted as X influences Y from the context of the discussion, from the interviewer’s notes with causal diagram snippets, or from the stated causality in the transcript of the interview. More subtle references to causality were interpreted from the context of the interview, for example: “... references to quite short distance lightweight [freight] movements – it’s something we have had an awful lot of talk about over the last few years but it’s quite a big policy shift” is interpreted as an influence between transport and planning policy and the operation of last mile delivery services. Similarly, the dialog “I mean Sunderland are already sort of struggling in terms of being a poor relation to Newcastle so it’s just a fear amongst certain council members that the Leamside line could further devalue Sunderland’s position in the region …” is coded as an influence from the relationship between local authorities on the business case to re-open the Leamside Line.

Some generic observations were made during the coding process. First, on occasion, the stakeholder did not agree with a concept in the scenario and departed from looking at the causality between the elements of the scenario and instead challenged the detail of the concept. Their comments then focussed on the shortcomings in the concept and not on how the path to it may either succeed or fail. The interviewer was drawn into switching from the “what if” questions to the “why not” questions, albeit with the same causality goal. Secondly, the scenario was multi-faceted, which complicated the elicitation. However, had the scenario and interview been constrained to be too narrow, stakeholders would not have had the opportunity to investigate it in the wider context and the analysis would have been artificially limited. At the end of the coding exercise, an exercise was carried out to rationalise the variables using two techniques. The first used cluster analysis where the measure of similarity was the correspondence in the links to and from each variable. The second was addressed by manual review. Examples of the cluster analysis and the merging of variables include:

• The variables “Leamside Business Case” and “Leamside Re-opened”: Although very similar in their links, one is the business case made to justify the other and the presence of a strong business case for a transport development does not necessarily imply the decision will be made to fund it. These two variables were kept distinct. • The variables “Strong Political driver” and “Political Ambition and Foresight”: These are adjacent in the cluster analysis and, on examination, were seen to both refer to the leadership required to instigate and carry out a transport project. The former tended to be used to refer to institutional leadership, the latter to a project champion for a specific development but the comments made by the stakeholders were similar. These two variables were merged into one “Political Action Initiated” variable. • The variables “LA Collaboration” and “Parochial attitude of NE LAs”: The collaboration and the parochial attitude variables on examination are opposing descriptions of a similar attribute; the tendency of Local Authorities to collaborate or compete; and were combined into a single “LA Political (dis)Unity” variable. • The “Social Entitlement to Travel” variable refers to a need for free travel in poorer areas while “Social Justice” refers to planning for transport availability. Similarly “Concessionary Travel” refers to a subset of the population (the over 60s) and their existing entitlement to free travel. These three variables were merged to a single “Social Inclusion” variable.

4.3 Results Emerging from the Uncertainty Analysis The goal of the uncertainty analysis was to gain insight into the role of the variables as drivers in the decision process identified both by category and individually. The Influence-Dependency graph for the full set of variables for all eleven stakeholders is shown in Figure 5. Only the variables in the upper half of the space have been labelled for clarity. The complete set of variables, broken down by category, is shown in Figure 6. Some of the key variables are discussed here: Political Action Initiated: Conversations around this variable refer to the belief that the decision to undertake a development is primarily a political decision. Typical comments on institutional determination included: “Investment, political will and community interest all required in sustainability “, “Northern Powerhouse gives political will.” Typical comments on project ownership included: ”Needs a strong officer in the council”, “someone with the expertise of doing it has to come and lead it“, “Integration needs someone in Government, It needs to be brokered.”

Looking at other instances of transport planning and regional politics, Hensher et al. (2015), after exhaustive analysis of the optimum choice between light rail based trams and bus rapid transit discovered that often the “wrong” choice was made but that it was still successful. Their closing comment was: “What appeared to make a difference in these cases was the existence of a champion who drove through an implementation package that turned out to be successful”. In other regions of the UK in recent years the post of a directly elected Mayor has been created in English cities (UK Parliament, 2000) (Stevens, 2006) whose role it is to move decision making from a traditional committee-based system to an executive based model. In the case of the Mayor of the Greater Manchester Combined Authority (GMCA) area it is explicitly: “ to steer the work of Greater Manchester’s Combined Authority, leading on issues such as the economy, transport, police and fire services,” (GMCA, 2017). The findings of the Leamside Line case study emphasise the need for such direct political based leadership. LA Political (Dis)unity: The disunity between authorities is shown by comments such as: “we wouldn’t be happy with a service that goes out to Metro Centre and Team Valley but you can’t get it from Gateshead”, “because we are all quite parochial in the North East and local authorities don’t work together in that way” and “The North East is, in words only, moving towards a combined transport authority.” Reasons emerge for the rivalry: “City competition for jobs” and “financial reward by having houses occupied [in their area] in terms of how much council tax they will raise.” Comments in the interviews often drew comparison with Transport for Greater Manchester (TfGM) as an example of an agency which coordinates transport policy and planning over a regional area and hence is more successful in winning funding for major transport projects compared to less coordinated regions. Recent legislation in the UK Cities and Local Government Devolution Act (UK Govt., 2016) provides for the creation of sub-regional transport authorities (in effect supra-Local Authority) “to co-ordinate the carrying out of transport functions in relation to the area that are exercisable by different constituent authorities, with a view to improving the effectiveness and efficiency in the carrying out of those functions;” with one early example being the formation of TfWM (Transport for the West Midlands) to fulfil that role in Birmingham city and six other surrounding local authorities (WMCA, 2017). The same act also provides for the creation of the directly elected Mayoral posts to provide executive leadership. Economic and Employment Growth: This variable is one of the key variables driving uncertainty in the Leamside Line planning being both highly dependent, and influential. It is primarily concerned with the local economy of the North East and though it was noted by some interviewees that it is not dissimilar from the overall economy of the UK and the global economy, others commented that recession (specifically the 2008-10 crash) may be deeper and differently timed in the North East relative to other UK regions. Spending on national infrastructure during times of recession to promote economic growth is however one of the fundamental tenets of Keynesian economics (Keynes, 1936) and that doctrine would suggest that investment in the Leamside Line would be advised to promote growth and recover from recession. The view from the stakeholders however was that the Leamside Line would not be funded until economic recovery was underway and the politics of austerity in the 2010 and the 2015 UK Governments had passed. Coincidentally, soon after the stakeholder interviews were conducted, the UK infrastructure commission was created with a goal to invest more in UK infrastructure (HM Treasury, 2016) potentially changing the influences and dependencies on growth and on government spending as

11 well as making more funding available for UK infrastructure, including transport. That action, a reversal of policy during the term of the Leamside Line case study, only reinforces the finding that the role of economic growth and its associated policies and funding is a strong factor in defining the uncertainty in the decisions in the initiatives surrounding the Leamside Line. Other Variables: Finally in this review of a selection of variables, there are some predictable results. For example, the role of the Department for Transport ( DfT)and Highways England (H.E.)is to manage and control transport developments in the UK. The position of the “H.E. DfT Action” variable places these two organisations in what Vester (2012) describes as the neutral regulation zone of the Influence Dependency Diagram confirming that role in the analysis. The “Public Opinion” variable has a similar controlling role influencing politicians and authorities and being influenced by the results of their transport development decisions and is also found in the neutral regulation zone. The relatively low priority of the sustainability aspect of the Leamside Line however was unexpected considering the growing emphasis sustainability in transport planning. This exposes a discrepancy between the factors found to be important in the opinions of the cohort of stakeholders, and the factors deemed to be important in the common understanding of transport development priorities. It is worth noting that the prime influential drivers of the project; the political leadership and the unity (or lack of unity) between Local Authorities emerged from the stakeholder interviews and were not explicitly referenced in the normative scenario. These variables were created in response to stakeholder inputs, volunteered as they commented on the plausibility of the events leading to the scenario being realised in its target year of 2030. This emergent view of uncertainty demonstrates the advantage of the methodology based on open ended interviews and an inductive approach to coding them. 5. Conclusions

Transport planning is an uncertain process which in theory exists in a world of rational analysis and sound project selection, but in practice, it functions within a complex policy and infrastructure ecosystem. The goal of this research was to devise and trial a novel method to identify the key influences, and the causes of uncertainty in the early stages of a transport project by linking techniques to identify and manage that uncertainty derived from multiple disciplines, namely scenario planning, qualitative data analysis, and complexity theory - as embodied in a structural dynamics model. This paper reports the findings of the application of this novel integrated approach, applied to the case study of the Leamside Line in the North East of England. Since the completion of the scenario planning exercise in 2013-2014, evidence has emerged that substantiates the key findings of the analysis. The most influential variables, crucial to the inception of the Leamside Line project were found to be the ability of local authorities to present a united approach to project development and the presence of strong leadership. Interestingly, action in the UK, formalised in legislation in 2016 (UK Govt., 2016), provides for the creation of integrated transport authorities and executive leadership and research based in Helsinki (Neuvonena and Ache, 2017) using backcast scenario techniques to investigate future options for transport governance made similar observations on the benefits of local authority collaboration. Furthermore, in the same period, a policy reversal was executed in the UK to move away from austerity and towards investment in infrastructure with the creation of the Infrastructure Commission (HM Treasury, 2016) mirroring the case study finding that the economic environment

12 formed a critical uncertainty. This comparability of the findings in the Leamside Line case study with those in actual developments elsewhere in the UK lends credibility to the results of the case study and shows that these influential drivers of the project are not unique to the Leamside Line project but have been observed elsewhere and solutions found. It should be noted that these developments, found to be key to project initiation in the case study, were not mentioned in the back cast scenario and emerged in the elicitation stage. This in turn demonstrates the value of the inductive coding system adopted in the data analysis. One of the findings of the case study presented here was that economic policy provided one of the major uncertainties in the project’s planning stage. Also, the position of the variable describing the actions of the UK Department for Transport and Highways England, is in the neutral regulation zone of the Influence– Dependency diagram described by Vester (2012) which matches the stated role of these two organisations which is to manage and control transport and road network developments. An unexpected result was the relatively low priority of the sustainability aspects of the Leamside Line, despite the fact that stakeholders with professional and political interests in sustainability had been included in the cohort. This exposes an inconsistency between the factors found to be important in the opinions of the cohort of stakeholders, and the factors deemed to be important in the common understanding of transport development priorities. This finding demonstrates the power of the methodological and analytical approach developed here, and its ability to step out from conventional understanding of transport planning uncertainty. The methodology developed here was not derived from the conventional transport modelling used in assessment, repurposing that model to perform uncertainty analysis. Instead the methodology was designed explicitly to identify the drivers of uncertainty in the transport planning and decision making process and eschewed the need to contribute to a “Go / No Go” decision. The method is designed to identify the factors of importance in decomposing and understanding a “wicked problem” (Rittel and Weber, 1973) and subsequent to this research, Wright et al. (2018) developed a six point framework to evaluate the capability of methods designed to assess “wicked problems”. In this framework, the method developed in this research: 1) Has a sound theoretical underpinning through prior work in the structural dynamics method by Godet (2011) and Vester(2012). 2) Is complete in that it is designed to readily widen to encompass the scope of the problem and not to constrain the study within preordained limits. This is demonstrated by the emergence, during elicitation, of aspects of the case study not included in the initial scoping. 3) Has a broad constituency, made explicit in the required feedback loop in which the range of stakeholders must be assessed after initial analysis of the findings. 4) Encompasses multicultural values, albeit subject to stakeholder selection. 5) Is explicitly designed to uncover uncertainty. However, with regards to Wright et al.’s (2018) sixth point- identifying an aim to seek a resolution of the problem. The method described here ends with an understanding of both the problem and the influences and uncertainties contained within it. It does not however prescribe actions, these must come from the active stakeholders based on the insights offered by this analysis.

5.1 Transferability In order to apply this method, the project to be studied should show significant uncertainty; such that its progress is either being blocked or its decision making process comes into question. In the case study used in this research, several studies had been undertaken to explore options to bring a valuable asset back into use, but no clear strategy emerged and no actions were planned. This was despite a widespread intention being evident that something should be done. There is, however, nothing unique to transport planning about these types of issues; the same can be stated concerning land use, infrastructure, economics, environmental management, product development, marketing, and politics. The requirements for application of this methodology are solely that the issue to be analysed must be identifiable, capable of being subjected to an initial scoping exercise, and that there must be an organisation or person sufficiently engaged to commission the application of this methodology. The results themselves derived from this case study however, should not be regarded as transferable. This research was not intended to identify generic uncertainties in transport planning. If similar analysis was carried out in another context, perhaps where an executive mayor and a unified transport authority were well established and in a different economic environment, then different uncertainties and influences would doubtless emerge. Instead, the goal of this research was to devise and trial the methodology, and while the results of the case study are both interesting and highly plausible, their purpose in this research is not to pronounce facts on uncertainty in transport planning per se, but to comment on the results derived from the methodology for this particular case study. To develop this methodology further, ideally, more case studies would be conducted with a goal to improve the techniques described here. If studies were to be undertaken on a live project, then post study workshops, attended by stakeholders could be held, the analysis presented and collectively acted upon. The critical evaluation factor would then become: Were stakeholders, on examining the findings of this methodology, empowered to act to change their views, the views of others, or the organisational factors that were shown to contribute to the uncertainty in the project, and, if so, did this then allow the project to progress?

In which case, we could be satisfied that the methodology would have achieved its goal.

5.2 Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. 6. References

AECom. (2006) 'Tees Tyne Connectivity Study Final Report', [Online]. Available at: www.strategyintegrationne.co.uk/displaypagedoc.asp?id=644 (Accessed: 25/09/2012).

AECom. (2007) Leamside Final Report (V3), November 2007 in Nexus.

AECom. (2010) 'Rail Freight Review 2010', [Online]. Available at: http://www.tyneandwearfreight.info/rail_freight/pdf/Rail Freight Review 2010.pdf. (Accessed: 25/09/2012).

Amaya Moreno, L., Defterli, O., Fügenschuh, A. and Weber, G. (2014) 'Vester’s Sensitivity Model for Genetic Networks with Time-Discrete Dynamics', in Lecture Notes in Computer Science. Vol. 8542.

ATOC. (2009) 'Connecting Communities Expanding Access to the Rail Network', [Online]. Available at: www.atoc.org/clientfiles/.../ConnectingCommunitiesReport_S10.pdf (Accessed: 18/09/2012).

Banister, D. and Berechman, J. (2000) Transport Investment and Economic Development. Routledge.

Bell, M. C. (2013) SECURE (SElf Conserving URban Environments). Available at: https://www.secure- project.org/ (Accessed: 17/05/2013).

Berechman, J. (2009) The Evaluation of Transportation Investment Projects. Routledge.

Booker, J. M. and McNamara, L. L. (2004) 'Solving black box computation problems using expert knowledge theory and methods', Reliability Engineering and System Safety 85, pp. 331–340.

Byrne, D. and Callaghan, G. (2014) Complexity Theory and the Social Sciences New York: Routledge.

Cacuci, D., Ionescu-Bujor, M. and Navon, I. (2005) Sensitivity and Uncertainty Analysis: Applications to large-scale systems. Boca Raton: Chapman & Hall.

Cairney, P. (2012) 'Complexity Theory in Political Science and Public Policy', Political Studies Review, 10, pp. 346-358.

Chakraborty, A. (2011) 'Enhancing the role of participatory scenario planning processes: Lessons from Reality Check exercises ', Futures, 43, (4).

Chermack, T. J., Lynham, S. A. and Ruona, W. E. A. (2001) 'A review of Scenario Planning Literature', Future Research Quarterly, Summer 2001, pp. 7-31.

Colander, D. and Kupers, R. (2014) Complexity and the Art of Public Policy: Solving Society's Problems from the Bottom up. Princeton: Princeton University Press.

Cole, A. (2006) 'The Influence Matrix Methodology: A Technical Report', [Online]. Available at: http://icm.landcareresearch.co.nz/knowledgebase/publications/public/iMatrix_Tech_Repor t.pdf (Accessed: 05/09/2015).

Courtney, H. (2001) 20/20 Foresight: Crafting Strategy in an Uncertain World Harvard Business School Press.

de Jong, G., Daly, A., Pieters, M., Miller, S., Plasmeijer, R. and Hofman, F. (2007) 'Uncertainty in traffic forecasts: literature review and new results for The Netherlands', Transportation, 34, (4), pp. 375-395.

DfT. (2009) 'Guidance on Local Transport Plans', [Online]. Available at: http://webarchive.nationalarchives.gov.uk/tna/20100210180753/http://www.dft.gov.uk/pg r/regional/ltp/guidance/localtransportsplans/ (Accessed: 08/07/2012).

DfT (2016) Transport analysis guidance: WebTAG. Available at: https://www.gov.uk/guidance/transport-analysis-guidance-webtag (Accessed: 19/02/2017).

Divall, C. (2017) 'High speed 2 where? A historical perspective on the ‘strategic case’ for HS2', in Transport Policy: Learning Lessons From History. pp. 97-114.

Dreborg, H. K. (1996) 'Essence of Backcasting', Futures, 28, (9), pp. 813-828.

Durham CC. (2010) 'Local Transport Plan 3', [Online]. Available at: http://www.durham.gov.uk/Pages/Service.aspx?ServiceId=5685 (Accessed: 30/05/2012).

Flood, R. L. and Carson, E. R. (1993) Dealing with Complexity. An Introduction to the Theory and Application of Systems Science.2nd ed New York: Plenum.

Funtowicz, S. and Ravetz, J. (1990) Uncertainty and Quality in Science for Policy. Dordrecht: Kluwer.

Giaoutzi, M., Stratigea, A., van Leeuwen, E. and Nijkamp, P. (2011) 'Scenario Analysis as a Foresight tool in Agriculture', Unpublished, In preparation.

GMCA (2017) Mayor of Greater Manchester: Andy Burnham. Available at: https://www.greatermanchester-ca.gov.uk/themayor (Accessed: 02/01/2018).

Godet, M. (2000) 'The Art of Scenarios and Strategic Planning: Tools and Pitfalls. ', Technological Forecasting and Social Change 65, pp. 3-22.

Godet, M. (2011) Methods of prospective: Micmac: Structural Analysis. Available at: http://en.laprospective.fr/methods-of-prospective/softwares/59-micmac.html (Accessed: 31/10/2011).

Godet, M., Durance, P. and Gerber, A. (2009) Strategic Foresight: La Prospective: Use and Misuse of Scenario Building. [Online]. Available at: http://en.laprospective.fr/dyn/anglais/ouvrages/sr10veng.pdf (Accessed: 05/09/2015).

Gordon, T. J. (1968) 'Initial Experiments with the Cross-Impact Matrix Method of Forecasting', Futures, 1, (2), pp. 100-116.

Grekos, M. (2014) 'Airports commission: Interim report', Journal of Planning and Environment Law, (4), pp. 420-422.

Harries, C. (2003) 'Correspondence to what? Coherence to what? What is good scenario-based decision making?', Technological Forecasting & Social Change, 70, pp. 797-817.

Hensher, D. A., Ho, C. and Mulley, C. (2015) 'Identifying preferences for public transport investments under a constrained budget', Transportation Research Part A, 72, pp. 27-46.

Hickman, R., Ashiri, O. and Banister, D. (2009) ' Achieving carbon efficient transport: backcasting from London.', Transportation Research Record, 2139, pp. 172-182.

HM Treasury (2016) Infrastructure at heart of Spending Review as Chancellor launches National Infrastructure Commission. Available at: https://www.gov.uk/government/news/infrastructure-at-heart-of-spending-review-as- chancellor-launches-national-infrastructure-commission (Accessed: 02/01/2018).

Holland, J. H. (2014) Complexity. A Very Short Introduction. Oxford: OUP.

Keynes, J. M. (1936) The General Theory of Employment, Interest, and Money. London: McMillan.

Kikuchi, S. (2005) 'Study of Transportation and Uncertainty', in Attoh-Okine, N. O. and Ayyub, B. M.(eds) Applied Research in Uncertainty Modeling and Analysis. Springer, pp. 303-319.

Lewins, A. and Silver, C. (2008) Using Software in Qualitative Research. SAGE.

Lindblom, C. (1959) 'The Science of Muddling Through', Public Administration Review 19, (2), pp. 79- 88.

Lindblom, C. (1979) 'Still Muddling, Not Yet Through', Public Administration Review, , 39, (6), pp. 517 - 526.

Lindgren, M. and Bandhold, H. (2009) Scenario Planning . UK: Palgrave MacMillan.

Marchais-Roubelat, A. and Roubelat, F. (2008) 'Designing action based scenarios', Futures, 40, (1), pp. 25-33.

Marchau, V. A. W. J. and van der Heijden, R. E. C. M. (2003) 'Innovative methodologies for exploring the future of automated vehicle guidance', Journal of Forecasting, 22, (2-3), pp. 257-276.

Marsden, G. and Reardon, L. (2017) 'Questions of governance: Rethinking the study of transportation policy', Transportation Research Part A, 101, pp. 238-251.

Mattot, L. S., Babendrier, J. E. and Purucker, S. T. (2009) 'Evaluating uncertainty in integrated environmental models', Water Resources Research, 45.

Morgan, M. G. and Henrion, M. (1990) UNCERTAINTY. A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis. Cambridge University Press.

Morris, M., D. (1991) 'Factorial Sampling Plans for Preliminary Computational Experiments', Technometrics, 33, (2), pp. 161-174.

Muric, G., Macura, D., Gospic. N and Bogojevic, D. (2013) An Approach to Assess Criticality of Elements in the Process of Information Infrastructure Protection., in Telsiks 2013 11th International Conference on Telecommunication in Modern Satellite, Cable and Broadcasting Services, Nis, Serbia.

Network Rail. (2007) 'Freight Route Utilisation Strategy ( Archived)', [Online]. Available at: http://archive.nr.co.uk/aspx/4449.aspx (Accessed: 25/03/2018).

Network Rail. (2010) 'East Coast Main Line 2016 Capacity Review', [Online]. Available at: ttp://www.networkrail.co.uk/browse documents/rus documents/route utilisation strategies/east coast main line/east coast main line 2016 capacity review/east coast main line 2016 capacity review.pdf (Accessed: 01/10/2012).

Neuvonena, A. and Ache, P. (2017) 'Metropolitan vision making – using backcasting as a strategic learning process to shape metropolitan futures', Futures, 86, pp. 73-83.

Nijkamp, P., Rienstra, S. A. and Vleugel, J. M. (1998) Transportation Planning and the Future. Wiley.

Ortuzar, J. D. and Willumsen, L. G. (2011) Modelling Transport (4th Edition) 4 ed London: Wiley.

Peterson, G. D., Cummimg, G. S. and Carpenter, S. R. (2003) 'Scenario Planning: a Tool for Conservation in an Uncertain World', Conservation Biology 17, (2), pp. 358–366.

Porter, M. (1980) Competitive Strategy.: Macmillan.

Rasouli, S. and Timmermans, H. (2012) 'Uncertainty in travel demand forecasting models: literature review and research agenda', Transportation Letters, 4, (1), pp. 55-73.

Refsgaard, J. C., van der Sluijs, J. P., Hojberg, A. L. and Vanrolleghem, P. (2005) 'Hamoni-CA Guidance: Uncertainty Analysis', pp. 46. [Online]. Available at: www.harmoni- ca.info/toolbox/Model_Uncertainty/index.php (Accessed: 10 April 2012).

Rittel, H. W. J. and Webber, M. M. (1973) 'Dilemmas in a General Theory of Planning ', Policy Sciences, 4, pp. 155-169.

Robinson, J. (2003) 'Future subjunctive: backcasting as social learning', Futures, 35, (8), pp. 839-856.

Rodier, C. J. (2007) 'Beyond Uncertainty: Modeling Transportation, Land Use, and Air Quality in Planning', [Online]. Available at: http://transweb.sjsu.edu/mtiportal/research/publications/summary/0701.html (Accessed: 30/08/2011).

Rowe, W. D. (2001) 'Managing Uncertainty', [Online]. Available at: http://rowe- res.com/Managing%20Uncertainty.pdf (Accessed: 27/03/2011).

Saltelli, A., Chan, K. and Scott, E. M. (2008) Sensitivity Analysis. Wiley.

Scolobig, A. and Lilliestam, J. (2016) 'Comparing Approaches for the Integration of Stakeholder Perspectives in Environmental Decision Making', Resources, 5, (4).

Soria-Lara, J. A. and Banister, D. (2017) 'Dynamic participation processes for policy packaging in transport backcasting studies', Transport Policy, 58, pp. 19-30.

Stevens, A. (2006) White Paper proposes stronger mayors and more power to English communities. Available at: http://www.citymayors.com/politics/uk_whitepaper06.html (Accessed: 02/01/2018). Sykes, P., Bell, M. C., & Dissanayake, D. (2018). "Combined use of a backcast scenario and cross- impact matrix analysis to identify causes of uncertainty in a nascent transport infrastructure project. " Transportation Research Part B. Methodological, 116, 124–140. Transport Scotland (2017) Aberdeen Western Peripheral Route / Balmedie to Tipperty. Available at: https://www.transport.gov.scot/projects/aberdeen-western-peripheral-route-balmedie-to- tipperty/ (Accessed: 15/08/2017).

Tyne and Wear Integrated Travel Authority. (2011) 'Keep Tyne and Wear Moving LTP3: The Third Local Transport Plan for Tyne and Wear Strategy 2011 - 2021', [Online]. Available at: http://www.tyneandwearltp.gov.uk/ (Accessed: 30/05/2012).

UK Govt. (2016) Cities and Local Government Devolution Act 2016. Available at: http://www.legislation.gov.uk/ukpga/2016/1/pdfs/ukpga_20160001_en.pdf (Accessed: 25/09/2018).

UK Parliament. (2000) Local Government Act 2000 http://www.legislation.gov.uk/ukpga/2000/22/contents. van der Heijden, K. (1996) Scenarios, the art of strategic conversation. Wiley. van Geenhuizen, M. and Thissen, W. (2007) 'A Framework for Identifying and Qualifying Uncertainty in Policy Making: The Case of Intelligent Transport Systems.', in van Geenhuizen, M.(ed), Policy Analysis of Transport Networks. Ashgate.

Vester, F. (2012) The Art of Interconnected Thinking: Tools and concepts for a new approach to tackling complexity.2nd ed Munich: MCB.

Vigar, G. (2017) 'The four knowledges of transport planning: Enacting a more communicative, trans- disciplinary policy and decision-making', Transport Policy, 58, pp. 39-45.

Wachs, M. (1985) 'Planning, Organizations and Decision Making: A Research Agenda', Transportation Research A, 19A, (5/6), pp. 521-531.

Walker, W. E., Harremoes, P., Rotmans, J., Van der Sluijs, J. P., Van Asselt, M. B. A., Janssen, P. and Krayer Von Krauss, M. P. (2003) 'A Conceptual Basis for Uncertainty Management in Model- Based Decision Support ', Integrated Assessment, 4, (1).

Weaver, W. (1948) 'Science and Complexity', American Science, 36, pp. 536-544.

Welde, M., Eliasson, J., Odeck, J. and Borjesson, M. (2013) 'The Use of Cost-Benefit Analyses in Norway and Sweden: A Comparison', European Transport Conference. Frankfurt, AET, pp. 1- 13.

WMCA (2017) Transport for the West Midlands. Available at: https://www.tfwm.org.uk/ (Accessed: 12/07/2016).

Worsley, T. and Mackie, P. (2015) 'Transport Policy, Appraisal and Decision-Making', [Online]. Available at:

http://www.racfoundation.org/assets/rac_foundation/content/downloadables/Transport_p olicy_appraisal_decision_making_worsley_mackie_May_2015_final_report.pdf (Accessed: 12/06/2016).

Wright, G., van Der Heijden, K., Burt, G., Bradfield, R. and Cairns, G. (2008) 'Scenario planning interventions in organizations: An analysis of the causes of success and failure', Futures, 40, pp. 218-236.

Wright, G., Cairns, G., O'Brien, F. A. and Goodwin, P. (2018) 'Scenario analysis to support decision making in addressing wicked problems: Pitfalls and potential', European Journal of Operational Research.

Figures

Figure 1 Influence Dependence Space (After Vester (2012))

Sample Causality Matrix V1 V2 V3 V4 V5 V6 V1 x xx V2 xxx x V3 x V4 x x V5 V6 xx x

Figure 2 Sample Causality Network

Figure 3 Influence Chain for variable-1

Figure 4 Summary of Integrated Method

Ref Job Description Employer Type Relevant Knowledge #1 Urban Planner Local Authority Leamside Line is within job remit. #2 Spatial Policy Team Leader Local Authority Leamside Line is within job remit. #3 Accessibility Consultant Consultant Extensive knowledge of public transport development practice. #4 Strategy Consultant Consultant Strategic advice in transport policy development. #5 Planners (2) Local Authority Leamside Line is within job remit. #6 Sustainability Manager Local Authority Relevant knowledge from adjacent Local Authority. #7 Planner Local Authority Leamside Line is within job remit. #8 Regional Director Consultant 20 years in transport planning consultancy. #9 Sustainable Transport Pressure group Lobbies for sustainable transport Activist development in the NE. #10 Consulting Director National Advisory Director of transport technology Body related professional services. #11 Transport Manager Local Authority Relies on transport provision for job function. Table 1 Stakeholder Summary

Figure 5 Complete Set of Variables

Figure 6 Variables broken down by Category

Appendix A: The Leamside line in 2030

Part of the network In 2030 the Leamside line is an integral part of the transport network in NE England. The public transport network consists of the Tyne metro, the bus network, the Tees valley metro (recently launched and built from the pre-existing network of suburban rail services) and medium distance rail between Middlesbrough and Newcastle based on the Leamside and Stillington lines which were re- opened in 2025. Coastal train services also run via Sunderland. The opening of the Leamside line allowed more freight movements south of Newcastle, Rail freight in 2030 also now includes quite short distance lightweight movements within the region as well as the long distance more traditional heavy bulk movements. The freight centre at Tursdale is a key local hub linking the rail and road network.

Public Transport Public Transport in the NE has been rising since 2000. In the first decade of the century, the rail franchises that were let on the basis of no growth in patronage saw on average a 30% growth in the first 10 years, The NE was starting a slow move back to public transport that was continued through into the 2020s. The Adonis report in 2013 on NE development made choices stark: ring Newcastle with motorways, hope for technology solutions to vehicle emissions and continue to manage congestion; or make the investment in high quality public transport and reduce car use. The choice fell on the side of public transport development allied to considered congestion mitigation in urban and trunk road hot spots. This strategy worked well with both the EU drive to remove fossil fuelled cars from urban areas by 2050 and the continued problem of range anxiety of electric vehicles which still precluded their use for long distances. The NE public transport network is heavily used both for convenience and because it is also cheaper than running and parking a car. The commuter service on the Leamside line consists of a light trains running between Middlesbrough and Newcastle. The train stops at Stockton, Thornaby, Stillington New Town, East Durham Parkway (previously known as Belmont Park and Ride), Fencehouses New town, and Washington before arriving in Newcastle Central where it interchanges with the Metro. Journey time is around 1 hour and the ticket is part of the NE England integrated travel card, so for regular commuters, ticket price is not an issue. Stations are well lit, safe places with good connections to the road network and to the local bus services. East Durham Parkway is a good example; the platform, up on the rail embankment, is connected by covered walkways down to the road level where electric buses run a shuttle service into Durham every few minutes (The buses are flash charged at the terminus as they wait for a few minutes before making the return journey). The A1 regional motorway is just 200 yds away from the carpark and there is extensive car parking with electric vehicle recharge points.

New Towns Living in Fencehouses The new rail link was a catalyst for new development at Fencehouses on the Leamside Helen and Stuart and their children are a line. The town was designed as an exemplar of typical family in Fencehouses. Stuart works in low emissions practice with localised food the local utility company office and cycles to production waste management and work, Helen works in Middlesbro’ and takes employment. The rail link with frequent shuttle the train. Depending on the weather she buses running round the town was a key part of either takes the metro from the station to that design. Walking and Cycling is the main work or cycles. She prefers to use her own local transport mode with shops, schools and bicycle so leaves one in secure storage at the offices integrated within the residential areas station although the train carries bikes and rather than set aside on retail or office parks. “Boris bikes” are available. Their eldest Many people living in Fencehouses also work daughter, a student in Newcastle has a very there and commuting distances are short. In fact similar commute in the opposite direction. as the cost of travel is low, disposable income is Their son walks to school but his sports higher and the town is actually quite activities mean he, and his mates, need prosperous. The Leamside line is however a key transported to rugby and cricket pitches all link to Newcastle and Middlesbrough so using over the county. the train for leisure trips as well as commuting Like most parents, they find transporting to work and to tertiary education is a fact of life children to activities from music lessons to in Fencehouses. rugby training requires precision timing so, as they use their car, they do appreciate the Freight reduced congestion on the roads compared to The case to reopen the Leamside line was partly what their own parents had to cope with. built on redirection of heavy freight trains from the ECML to the Leamside line freeing up ECML capacity in the Durham – Newcastle pinch-point. The main heavy rail traffic is from Tyneport docks onto the national freight network and from the Nissan car plant to the docks. The new freight centre at Tursdale also provides an interface to the road network and onto the local freight operation. The local freight operation is what sets the Leamside line and the rest of the NE rail network apart from the conventional heavy rail freight operation. The key difference is that the distance over which rail is viable Freight has been dramatically reduced. The concept is close to The nineteenth rail freight depot in Tyne and being a “freight tram” with rapid automated handling Wear has just been opened. Its at Benton, close to the metro station and uses the metro of pallet sized freight, the use of the suburban rail lines. Freight cars pull into the siding and network and the metro, with small depots serving local automated loading and unloading systems areas and electric vans making the “last mile” delivery. manage micro and nano containers (3 and 1 The system already complies with the 2011 EU plans to cubic metres respectively) feeding them remove fossil fuel powered vehicles from urban straight into waiting vans for local delivery. centres and has helped boost the concept of urban Containers may be full of market garden market gardening pioneered at Fencehouses by produce, from Fencehouses, waste for recycling, or may be full of internet shopping making short distance low volume freight commercially orders about to be delivered over the final and environmentally more viable. mile by a fleet of electric vans The road to rail By 2013 there were several reports written on the future of the Leamside line, these discussed local 27 commuter operations, national rail operations and freight only. They had one thing in common; a conclusion that on their relatively narrow individual focus and with economic assessment based predominantly on travel time savings, the re-opening was hard to justify and the Leamside line could be summed up as a solution looking for a problem. The impetus to re-open came from several sources: the observation that rail passenger numbers were rising beyond what had been forecast causing a re-assessment of the economic case, including the value of emissions savings, including the value of economic growth due to urban agglomeration, and measuring social justice. These all improved the business case as did the drive to find infrastructure investment projects as the UK came out of recession. The fact that the Leamside line was still nominally an operational railway meant the legal process to re-open it was greatly simplified and did not block progress. The economic case for re-opening the Leamside in tandem with the Stillington line was made on the release of capacity on the East Coast Main line as long distance freight was diverted, on the forecast of mode shift from cars for local commuting between Newcastle, Washington, Durham (East Durham Parkway) and Middlesbrough. There was also support from Highways England as part of a programme of congestion management and traffic reduction on the regional roads. The case was made for the 2019 rail control period and construction started in 2022. Railway Timetabling The technical barriers to the re-opening, i.e. capacity John is a rail manager for the NE of England constraints at Middlesbrough station and management with responsibility for the routes south of of mixed use freight and light rail were overcome with Newcastle. new rail management systems. More logistical barriers, The priorities have changed. In 2010, flighting i.e. duplication of existing Metro, bus and suburban rail was necessary in the timetable to batch slow services were overcome through collaboration with the freight and express passenger trains on the bus companies to operate the new local rail service and line. The number of slower commuter trains to complement it with linked bus and metro services stopping at Chester le Street had also been with a common ticketing system. The significant reduced to improve throughput on the line. engineering problems were expensive (i.e. re-routing When the Leamside line re-opened in 2025, past the Victoria viaduct which remains a cycle route) some freight was routed on to it to relieve the and replacing collapsed embankments; but funding was bottleneck south of Newcastle. More freight made available. could be shifted, and as flighting was no The knock on effects of the new rail lines were several. longer necessary, passenger services could be With pressure relieved on the ECML, more commuter more evenly spaced. trains between Newcastle and Durham were able to Now in 2030, his problems are back. The stop at Chester le Street promoting more growth in rail Leamside line with its mixture of slow heavy use, less cars on the road and playing apart in the freight and light commuter trains is at growth of Chester le Street as a commuter town. The capacity. The growth in freight also means the requirement for more housing in the NE was not able ECML is back at full capacity, the extra to be met by additional growth in existing towns and commuter services stopping at Chester le Fencehouses was selected as the site for a new town Street are heavily used and he is under designed as an emissions free community. This, and the pressure to find more capacity on the need to reduce emissions from delivery vehicles in network. cities, was the catalyst for lightweight freight on the NE railways and new rail management systems.