Implementing bus rapid transit: A tale of two Indian cities

Research in Transportation Economics xxx (2014) 1e11

Contents lists available at ScienceDirect

Research in Transportation Economics journal homepage: www.elsevier.com/locate/retrec

Implementing bus rapid transit: A tale of two Indian cities Andrea Rizvi*, Elliott Sclar Department of Urban Planning, Columbia University, New York, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Available online xxx

The ongoing debate within the Bus Rapid Transit (BRT) community over the relative importance of 'appropriate' design standards, the 'right' institutional setup and the need for 'political will' to the success of projects obscures the larger importance of the planning process to outcome effectiveness. Political leadership, institutions and design are important conditions that must be considered in the context of one another, but they are also conditions that will change and be influenced by the planning process. Drawing on case studies of Janmarg BRT in Ahmedabad, and the Delhi BRT in India's capital, we demonstrate the role of the planning process in influencing BRT project outcomes. The planning process is too often viewed as a sequence of steps in which design, institutions and leadership provide an unchanging framework in which planning proceeds. This ‘one-dimensional’ view needs be re-framed in ‘three-dimensions.’ It must explicitly also consider approach (i.e. strategy and tactics) and timing (i.e. both moment of action and duration). These in turn reshape design, institutions and leadership. Findings suggest that such a 'three-dimensional' planning process, when well timed, incremental and pragmatic may help to overcome institutional and design weaknesses, and to solidify political support. This improves viability and long-term system sustainability. © 2014 Elsevier Ltd. All rights reserved.

Keywords: BRT Bus rapid transit Implementation Developing countries India

1. Introduction Bus Rapid Transit (BRT) has been promoted as a low cost panacea to problems of urban mobility (Hesse, 2010; Kubala & Barton, 2003; Wright, 2010). BRT holds the promise of myriad benefits e offering affordable service for the poor; the prospect of environmental improvement as well as much needed urban congestion relief (Currie, 2005; Hensher & Golob, 2008; Hidalgo & Gutierrez, 2013; Levinson, Zimmerman, Clinger, & Rutherford, 2002). BRT technology is particularly suitable for developing countries due to it's low capital costs, flexibility, and potential for integration with much-needed non-motorized transport (NMT) facilities (Fouracre & Dunkerley, 2003; Hidalgo & Graftieaux, 2008; Polzin & Baltes, 2002). It is not surprising that countries like India have enthusiastically adopted BRT systems, rolling out the technology to Delhi in 2008 and Ahmedabad in 2009. Today there are 6 operational BRT systems in India 1 with a further 13 currently being planned or under construction.

* Corresponding author. Department of Urban Planning, Columbia University, 1172 Amsterdam Ave., New York, NY 10027, USA. Tel.: þ1 212 854 3510. E-mail addresses: [email protected], [email protected] (A. Rizvi). 1 Operational BRTs include Pune (2007), Ahmedabad (2009), Delhi (2008), Jaipur (2010), Rajkot (2012), Indore (2013).

The Delhi BRT was one of the nation's first ‘full service’ BRT. Ten years in the making, the project has clear institutional responsibilities, significant political support and a thoughtful and inclusive design (DIMTS, 2010; Kishore, 2009; TRIPP, 2005b). And yet the Delhi BRT faced many operational challenges. It has been widely lambasted in the press as an abject failure, with public interest litigation leading to a court-ordered shut down of the system (Aaron, 2008; TH, 2008b). Today it continues to face public relations obstacles and despite high ridership has failed to be extended beyond its initial pilot length. In contrast the Janmarg BRT in Ahmedabad, launched little more than a year later, was successful in overcoming initial hiccups and is now internationally lauded as a success story, receiving multiple national and international awards.2 Implemented much faster than Delhi, it also benefited from a careful design, clearly assigned responsibilities, and strong political will. Today the system is well received by city residents, extending over 70 km while also recovering operating costs (CEPT, 2013, p. 12). What explains the difference in perceived outcomes of these two schemes? Global evaluations of BRT suggest a wide variety of factors that support success, including issues associated with

2 Awards include the Government of India's 2009 “Best Mass Rapid Transit System” award and the ITDP Sustainable Transport Award for 2011.

http://dx.doi.org/10.1016/j.retrec.2014.09.043 0739-8859/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

2

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

political leadership, institutions and design choices (Hesse, 2010; Hidalgo & Carrrigan, 2010; Hidalgo & Graftieaux, 2008; Kittelson and Associates & Consultants, 2007; Miller & Buckley, 2001; Wright & Hook, 2007). The latter has been the subject of extended recent debate within the BRT community developing ‘bronze’, ‘silver’ and ‘gold’ design and service standards (ITDP, 2012). While these factors are important, they overlook the role of planners and ‘good’ planning processes in influencing outcomes. This paper draws on cases of BRT in Delhi and Ahmedabad to show how planning processes impact both perceived and actual achievements, extracting lessons for India and beyond. 2. Background The rapid global expansion of BRT systems in recent decades has generated an equivalent expansion of literature on the nature, characteristics, advantages and impacts of this technology. (CFTE; DNA, 2010; Levinson, Zimmerman, Clinger, & Gast, 2003; Levinson et al., 2002; Polzin & Baltes, 2002; Vuchic, 2007, pp. 202e298). Strong BRT advocacy groups have also emerged. But despite its considerable promise, BRT systems have faced numerous obstacles during implementation. There has been considerable debate within the BRT community concerning the importance of design quality for successful BRT outcomes (ITDP, 2012). The role of institutional and political ‘success factors’ are also well documented (Gow & Morss, 1988; Hossain, 2006; Miller & Buckley, 2001; Thynell, Mohan, & Tiwari, 2010). But recent cross-country studies (Hidalgo & Carrrigan, 2010; Hidalgo, Custodio, & Graftieaux, 2007) also point to the possibility of improving outcomes through planning practice. These studies offer useful checklists for improving the planning process of BRT schemes. But while they serve as guidelines for BRT implementers they offer a one-dimensional understanding of the planning process as a ‘series of steps’,3 and do not reflect other critical dimensions of the process e most notably timing and planning approach. In the context of this paper, a broader three-dimensional definition of the planning process will be considered (refer Fig. 1). As shown in this schematic, the content of the planning process can still be understood as a sequence of events that commences at problem identification and concludes post-implementation. But planning process should equally be defined by the manner in which it is pursued e e.g. strategy and tactics employed; and its temporal qualities e e.g. duration, order and timing of events. The planning process thus includes not just the ‘what’, but also the ‘how’ and ‘when’ of the decision-making, technical analysis, citizen involvement and the negotiation involved in producing a plan. For the purposes of this study, the planning process is considered separately from institutional, political and design considerations. The intent is to show how the process itself, independent of these factors, can influence outcomes. In practice of course these distinctions are blurred as all factors interact closely in the production of any ‘plan’, ‘policy’ or ‘project’. To demonstrate the relevance of a three-dimensional framework, this paper describes the BRT planning process in Ahmedabad (2005 to present) and Delhi (1996 to present), comparing and contrasting experiences from initial conceptualization to operationalization. The discussion starts by examining city context, politics, institutions and design of the BRT systems, highlighting differences between the two cities. The cases then describe the content (steps), approach (strategy) and timing of the planning

3 Planning process is typically understood as a rational sequence of events starting with problem analysis, identification and evaluation of design alternatives, concept development, design, implementation and feedback.

Fig. 1. Three dimensional model of the planning process.

process and how it evolved in both cases, responding and contributing to the influence of politics, institutions and design, while also directly impacting outcomes. Information was collected from a variety of sources including published documents, meeting minutes, design and preparation documents, technical reports, newspaper articles, promotional pamphlets and other materials illuminating the underlying processes, events and stakeholders. In addition interviews were conducted with key decision makers, relevant agencies, officials and stakeholder representatives in India in September 2012. Meetings were also held with design teams of both systems, and visual inspections of the operating corridors undertaken. Follow-up interviews to verify data and probe more closely on key issues were conducted from January 2013 through to June 2013 using phone, skype and email. 3. The city, political and transport contexts of Delhi and Ahmedabad Delhi (formally known as the National Capital Territory of Delhi or NCTD4) is one of the largest metropolitan areas in India, incorporating the nation's capital e the city of New Delhi. The sprawling, poly-nucleated city serves a population of 16.8 million and is a key political, cultural and commercial center for the nation (Bose et al., 2001; GOI, 2011). Urban transportation in Delhi is road-based with mass transit provided by public and private bus service, with little attention to non-motorized transportation (NMT). Poor bus services has resulted in a decline in public transport ridership over the last decade that even the highly popular “Delhi Metro” has failed to stem. Private motor vehicle ownership in the city has increased rapidly, resulting in chronic congestion, deteriorating traffic safety and declining air quality in Delhi (DIMTS, 2010). Institutional governance in Delhi is complex, hierarchical and highly political, with a multiplicity of local, state and federal authorities engaged in the cities affairs leading to poor institutional clarity, duplication, delayed decision-making and serious coordination difficulties (Baijal, 2011).

4 The NCT has special status under the Indian Constitution, as neither an independent State nor centrally administered Territory. The elected Government of the NCTD (GNCTD) holds only limited state government powers.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

Ahmedabad is also a commercial and educational center but is much smaller and less complex city. Located in the fast-growing and rapidly urbanizing State of Gujarat, the city is relatively compact with a population of 7.2 million (GOI, 2011). Ahmedabad is heavily reliant on its congested but comprehensive ring radial road network. Average trip lengths in the city are short with a significant proportion of pedestrians and bicyclists. Motorization rates are low (but rising rapidly), dominated by more affordable two wheelers. Public transport ridership is lower than Delhi with users traditionally relying on overcrowded, unsafe and inefficient buses operated by Ahmedabad Municipal Transport System (AMTS), supplemented by private and state-run bus charter services. Rail lines infiltrate the city but serve long distance travel needs. Governance in Gujarat is highly political with a dominant middle class and strong religious divisions (Harriss, 2006), but over the last decade the state has been united under the powerful and longpresiding Chief Minister Narendra Modi. Institutional responsibilities for transport are clear and well coordinated across levels of government. 4. BRT design The design of the BRTs in Delhi and Ahmedabad (or “Janmarg”) BRT followed internationally accepted standards, with both systems comprising two, one-way, physically segregated central median bus lanes sited on existing road corridors with regularly spaced, ITS equipped, bus stations. Despite these physical similarities however there were distinct differences in design objectives and philosophies that had an important bearing on outcomes. Most fundamentally, the Delhi BRT is designed as an open system allowing all buses (and emergency vehicles) to freely enter and exit the bus corridor along the route (Kishore, 2009). This approach maximizes the catchment area (and thus commuters) without the need for transfers and feeder buses e essentially offering expedited service for existing bus routes and enabling them to compete against two-wheeler and auto-rickshaw service (Tiwari & Jain, 2012). In contrast, Ahmedabad implemented a closed system, providing a trunk service and customized fleet that operated exclusively on the dedicated lane separated from the main carriageway by railings (except at intersections/roundabouts). This ‘closed’ system was selected because it promised more control over service quality, as well as improving speed, reliability and efficiency of operation through off board ticketing and freedom from mixed traffic. Another important distinction between the two systems was the emphasis placed on NMT infrastructure. NMT was a primary motivator and assigned considerable priority in the design of the Delhi BRT e a dedicated high quality bicycle lane (2.5 m) and footpath (1.5 m) runs continuously on both sides of the corridor. This required a reduction in the width of existing motorized lanes. By removing buses from mixed traffic, the expectation was that overall traffic speed would increase (Kogdenko, 2011). Street lighting, landscaping, bicycle hire, vendor and parking facilities were also incorporated into the cross section (Tiwari & Jain, 2012). In Ahmedabad, NMT was assigned a lesser priority. Sidewalks for pedestrians (2.0 m), and bicycle lanes (2.0 m) (in some segments) were constructed along the vast majority of the bus corridors and signalized crosswalks were provided at junctions and stations to enable safe access, with grade separated facilities constructed at high demand locations (ADB, 2010; Tiwari & Jain, 2010). Other design differences occur in bus stations and operations. In Delhi parallel bus stations are spaced approximately 500 m apart located at near side of intersections to maximize boarding capacity and minimize signal timing (Ms. Chowdury, interview, September 18, 2013). Critics suggest that the location and sizing of these

3

stations has contributed to bus-bunching and congestion at intersections. This has been compounded by long delays caused by the six phase traffic signals required to manage the multiple traffic and bus movements as well as on-board manual ticketing (Baijal, 2011; Tiwari & Jain, 2012). Though accessibility concerns were paramount to Delhi designers, level access/egress is only partially implemented, restricted to the upgraded low floor bus fleet. In contrast, Janmarg stations are aesthetically distinctive, closed, offer dual side boarding and are spaced away from intersections at 800 m intervals (ADB, 2010). Each platform is built up to facilitate quick, safe and level boarding and is accessed by signalized pedestrian crosswalks (Kadri, 2010). The dedicated fleet that operates on the corridor is branded, with larger glass windows, wheelchair space, rods and handles for standing passengers, and single level low floor, wide entries, and improved seating. These modifications support larger flows and passenger mobility, particularly for the physically impaired. Engines and suspensions were updated to reduce emissions and provide a more comfortable ride. All buses are tracked and monitored linking to Passenger Information Systems (PIS) at stations (ADB, 2010; Kadri, 2010; NIUA, 2009, p. 48). 5. Institutional arrangements From its inception, the Ahmedabad BRT was a locally driven initiative e originated and designed in Ahmedabad and highly responsive to local conditions. Fig. 2 outlines institutional responsibilities for the project. The Ahmedabad Municipal Corporation (AMC), the key municipal agency in the development of the initial proposal for the BRT, remains the chief executing authority of the system (Tiwari & Jain, 2010). Concept design and implementation was led by the Center for Environmental Planning and Technology (CEPT) University in Ahmedabad with assistance and support from a variety of local consultants, as well as the New York based Institute for Transportation and Development Policy (ITDP). The use of predominantly local expertise not only ensured responsiveness to local conditions and technology transfer, but also was effective in keeping costs low (HIdalgo & Pai, 2009). This strong local ownership was matched at state and national levels with support provided through a steering committee under the State Urban Development and Housing Ministry; and financial and other incentives at the National level through the JnNURM program. Strong political commitment and consistent leadership characterized the project across levels of government (Thite, 2010; Tiwari & Jain, 2010). Institutional responsibilities were also clearly defined and aligned (Baijal, 2011). The Ahmedabad Janmarg Limited (AJL) was constituted as a special purpose vehicle (SPV) under the purview of AMC to serve as a dedicated and independent agency with operating authority (Pai & Hidalgo, 2009; Tiwari & Jain, 2010). ‘Ring-fencing’ of operations in this manner facilitated monitoring and oversight and thereby financial control of operations. It also allowed rapid and flexible decision-making. Importantly, the interests of AMC, AMTS and AJL were aligned through a shared Chairman/Chief Executive.5 Numerous tasks have been outsourced to the private sector to maximize expertise and efficiencies. Initial contractual hiccups (e.g. contract specifications, performance parameters) were overcome with trial and error that in some instances required multiple bidding of certain contract packages (Baijal, 2011). In contrast to Ahmedabad, institutional arrangements in Delhi are necessarily complex due to the multiplicity of federal, state and

5 The Municipal Commissioner of AMC is also the Chairman and Chief Executive of AJL.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

4

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

Fig. 2. Institutional responsibilities for Ahmedabad BRT.

local agencies that operate in the city6 (refer to Fig. 3). The multiple layers and tiered reporting structures meant that incentives of these different agencies were not aligned and not all fully supported the Delhi BRT. Nevertheless, institutional responsibilities for the Delhi BRT were relatively well defined e overseen by the Transport Department of the GNCTD, the project was managed by the Rail India Technical and Economic Services (RITES)7 with conceptual and technical advice provided by the Transportation Research and Injury Prevention Programme (TRIPP) within IIT Delhi India Institute of Technology (IIT) (DIMTS, 2010), supported by ITDP during the initial stages of conceptualization. In 2006, as design was finalized a new player was introduced. DIMTS e a special purpose vehicle (SPV) formed as a joint venture between the Government of Delhi and the Infrastructure Development and Finance Company (IDFC e a private company) e was entrusted with the implementation and operation of the BRT scheme. The Environmental Pollution Control Authority (EPCA) closely monitored progress (Kishore, 2009). Private sector partners were engaged for construction of the BRT corridor infrastructure, maintenance of corridor infrastructure, management of advertising, and provision of traffic marshals. Within the corridor, buses are operated by both DTC and licensed private operators. Though there was strong leadership from the Chief Secretary at different stages of the project, lines of communication between agencies tended to be formal and hierarchical (Baijal, 2011; Ponnaluri, 2011). There were also staff turnover at the state level impacting continuity and ownership of the process. The introduction of DIMTS provided impetus as construction started, but their late engagement in the process was initially problematic as they lacked insights from earlier decisionmaking processes (Baijal, 2011).

6 For example, in the case of the Delhi BRT corridor, the Public Works Department (PWD) of GNCTD owns the road; the Delhi Integrated Multimodal Transit System (DIMTS) has oversight of BRT lanes and bus stops; the Delhi Transport Corporation (DTC) owns the bus terminals and depots; the buses that ply the corridor are either owned by DTC or are private operators licensed by the Transport Department of the GNCTD and traffic enforcement falls under the federal jurisdiction of the Delhi Traffic Police. 7 RITES is a public sector undertaking of the Ministry of Railways.

This comparison suggests that the Ahmedabad BRT project benefited from a defined institutional structure, with aligned incentives and good interagency collaboration. Politically the project was supported at national, state and local levels and was based on thoughtful and locally based design. These were all important factors that contributed to project success. Though Delhi also had clear institutional responsibilities assigned for the project, the broader institutional environment was complex and made interagency coordination both essential and difficult. Institutional incentives were not aligned and political support was present but not consistent or continuous. Like Ahmedabad however, design in Delhi was thoughtful and local, based on the needs and priorities of the city. It seems that institutions, politics and design offer only a partial explanation of the quite varied outcomes in Delhi and Ahmedabad. The next section will show how the planning process also played a role. This is because the planning process not just responded to these initial conditions, but also served to influence political support, institutional strength and design acceptance, thereby making it's own singular contribution to project outcomes. 6. The planning process 6.1. Ahmedabad BRT The concept for a BRT in Ahmedabad arose as one of several parallel alternatives considered under a joint state/local comprehensive mobility study. It's primary objective was to reduce congestion and improve access in the city of Ahmedabad8 by offering a ‘new’ transportation system. BRT was identified as a preferred technology because the existing ring road structure largely matched the city's transport needs and could support the bus network. It was also flexible and expandable, could be implemented more rapidly than the metro alternative, and offered the

8 The study was jointly directed by the Gujarat Infrastructure Development Board (GIDB), AMC and AUDA and recommended a multi-pronged approach to congestion relief that included regional rail, metro, parking policies, improvements to the existing AMTS as well as provisions for a new citywide high-end BRT system.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

5

Fig. 3. Institutional responsibilities for Delhi BRT.

opportunity to improve public space and accessibility through provision of NMT facilities (NIUA, 2009, p. 48). It was a combination of these physical characteristics, a supportive national policy, clear objectives, careful timing and forward planning that led to the rapid implementation of BRT in Ahmedabad. Timing in particular was critical. Enthusiasm for the technology grew when Bogota Mayor Enrique Penalosa visited the city in November 2004 (Prof. Swamy, interview September 11, 2012). Concept development coincided with the “2005 Year of Urban Development” in the State of Gujarat, and with the city's “Accessible Ahmedabad” vision. At the national level it was supported by policy developments in the form of the emerging National Urban Transport Policy (NUTP) and its well-resourced, reform-linked investment program Jawaharlal Nehru National Urban Renewal Mission (JnNURM) (Iuchi, 2012). Foresight and preparation were also present in the form of Prof. Shivanand Swamy (Head of the Department of Transportation Planning at local university CEPT) who was recruited to assist AMC early in concept development and became a key actor in the implementation of the scheme. He was well versed in the technology, having worked on BRTS proposals since the early 1990s, guiding masters thesis topics on its application to Indian cities (Baijal, 2011). Widely supported, the conceptual design of the Ahmedabad BRT was initiated at an International Workshop in August, 2005 (Prof Swamy, interview, September 11, 2012). In June 2006 another high profile proponent of BRT, Shri I.P. Gautam was appointed

Commissioner of the AMC (Baijal, 2011). He personally advocated for the project at the National Ministry of Urban Development (MoUD), securing early support under the recently launched JnNURM program. By May of 2006, little more than one year after initial discussions, the Janmarg BRT received its first round of funding for the implementation of the initial 12 km demonstration corridor. This was followed in June 2006 by the remaining funding for Phase I of the scheme (total length 58 km) and a subsequent 30 km under Phase II (Pai & Hidalgo, 2009). Several important philosophies guided the design of the Ahmedabad BRT. Firstly, there was a defined and clearly stated effort to ‘connect busy places but avoid busy roads” (AMC, 2012). Routing was selected to connect destinations, but deliberately avoided congested roads. Secondly, there was an emphasis on ‘designing a network and not corridors’ to serve the whole (and not just part of the) city. Thirdly, the system emphasized equitable road space allocation amongst users and aimed to improve mobility for all by ensuring spatial coverage, employment coverage and improved accessibility (AMC, 2012) (Prof. Swamy, interview, September 11,2012). Driven by these philosophies, the initial master plan included 88 km of closed BRT corridors to be implemented in phases (NIUA, 2009, p. 48). There was a strong dose of pragmatism associated with the design process. The Detailed Project Report (DPR) is explicit about the precedence of ROW and ease of implementation over other more technical concerns in selecting initial corridors in phase 1.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

6

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

Pragmatism was also tempered by a keen political awareness of the need to consider all users. Though the aim was to provide very high quality bus service, the system was designed to ensure that ‘immediately after completion of the project, (all users would) receive same or better level of services” (AMC, 2007 pp. 6e4). Care was taken not to encroach on lane space of powerful motor vehicle users. The design process was iterative and drew heavily from other BRT experiences. Over 100 minor design modifications were made to avoid difficulties faced during implementation in Delhi and Pune. For example an initial decision to adopt curb-side bus stops (allowing for two bus stops on either side of the central lane) was altered as construction started based on feedback from additional design team trips to Bogota and Pereira. Instead a central island bus stop with dual sided boarding was adopted, saving costs and facilitating transfers (AMC, 2007). Similarly, learning from Delhi it was decided that in the interest of service quality control only BRT buses should be allowed to ply on the dedicated lane. In addition, bus stop locations were moved from junctions to improve flow, geometric standards for the road design were changed to take into account narrower ROW on some stretches, lighting was relocated from bus median to the curb, and widths of pedestrian and cycle paths were increased e all demonstrating the flexibility and receptiveness of the design team to feedback and adaptation (AMC, 2007; Baijal, 2011). An impressive feature of the Ahmedabad BRT was how rapidly it was implemented. In part this rapid construction (2 years e October 2007 to October 2009) was possible because of a collaborative institutional structure. The AMC led the decisions but included actors from state, city departments and academic institutions with CEPT playing a critical role in mediating and engaging stakeholders (Iuchi, 2012). But the strategic use of incremental construction also contributed to streamlined implementation. The first phase was constructed in 5 discrete sections, commencing with a 12.5 km pilot corridor (Baijal, 2011). Design for Phase II was not initiated until construction had commenced. This facilitated the iterative design process discussed above, allowing adjustments to be made based on feedback from Phase I. Another advantage of the incremental approach was that it helped to strengthen public support and maximize ridership. This was particularly evident in the initial 12 km demonstration corridor. Firstly, to overcome critics concerns about congestion, this initial corridor was strategically placed on a lower congestion route with moderate bus volumes (Kost, 2009). More difficult to implement corridors were delayed until later stages of implementation when initial gains had been solidified (Kost, 2009; NIUA, 2009, p. 48). This also allowed the system design to be tested and modified before moving onto more difficult routes. Additionally, trial rides on the demonstration corridor were offered to passengers free of charge. This familiarized commuters with the system and helped broaden public acceptance. It also ensured that customer feedback could be incorporated into operations before the system was officially launched and enabled testing of scheduling, bus operations and signal interface. Finally, concessions were made in operation during the initial phase. Despite its design as a closed system, buses were allowed to travel in mixed traffic outside the dedicated 12 km stretch, to maintain connectivity for passengers in the initial phase. After the demonstration corridor was completed the remainder of the first phase launched gradually in lengths of 5e10 km to maintain momentum (Prof. Swamy, interview, September 11, 2012) (NIUA, 2009, p. 48; Pai & Hidalgo, 2009). A number of proactive human resource actions taken during implementation also supported operational success of the Ahmedabad BRT. There was early recognition that enforcement would be critical to the effective implementation of the system, especially as

it required a significant change in travel behaviors amongst users (e.g. dedicated lane, pedestrian right-of-way etc.). To address this need Janmarg enlisted ex-servicemen to supplement traffic police, creating a dedicated, loyal and appropriately empowered cadre of professionals to oversee and trouble shoot operations on the ground (Dr. Katyal, interview, September 19, 2012). Janmarg drivers were carefully selected and received training for both physical and mental dimensions of their work. Practical and theoretical training was provided on docking procedures and safe comfortable driving techniques, supplemented by innovative yoga training to support mental and physical health (NIUA, 2009, p. 48). Janmarg adopted a policy of early, frequent and targeted community consultation and public relations efforts. Neither formally articulated, nor professionally developed the consultation process was distinguished by its transparency and active engagement of feedback. Comments from the public and special interest groups was actively sought, thoughtfully responded to, and captured in design and process modifications. The open approach toward the media in particular instilled a constructive interaction and was supplemented with written material. This philosophy of engagement and feedback continues today with administration and evaluation of monthly user surveys on the system (Prof Swamy, interview, September 11, 2012). Some of the key strategies employed during stakeholder consultations include:  Outreach started early in the process in the form of community workshops as part of the initial city mobility plans, and expanded to cover environmental, social, technology, and design issues with a broad range of stakeholders. Workshops served the purpose of both collecting feedback, but also to educate and allay public concerns regarding BRT proposals. AMC and Janmarg representatives worked with media outlets and international experts to promote BRT and it's projected benefits to city-dwellers. Strategic deployment of the Kankaria Lake loop segment to coincide with the 2009 carnival at the recently upgraded recreational lake was a promotional coup. Janmarg attracted positive publicity and was portrayed as a genuine public good, bringing revenues to the city and serving the daily needs of the people (AMC, 2007; Baijal, 2011)  Knowledge sharing was also critical to promoting learning and to building understanding and support amongst politicians, official and technical personnel. Representatives from other BRT cities were both visited and hosted to exchange knowledge on planning, implementation and operation of these systems. International workshops were also used strategically to engage expertize where required (Baijal, 2011).  Technology showcasing took the form of publicity and poster demonstrations in real estate shows and state conventions and was useful in broadening political support for the schemes. A prototype bus stop was also constructed to facilitate public understanding of the system and solicit feedback. But the masterstroke of the communication strategy was the introduction of three months of free trials to ‘showcase’ the new BRT technology in Ahmedabad. Not only did this allow potential passengers to test and understand systems and applications, but it also improved system performance by familiarizing drivers with corridor and docking practices, and gave Janmarg the ability to test and modify system operations prior to formal commencement of operations. Potential dissident groups were targeted during the trial period and given escorted rides to garner support for the system (Baijal, 2011; NIUA, 2009, p. 48) (Mr. Solanki, interview, September 13, 2012).  Branding. From its inception the Ahmedabad BRT image was carefully and consistently managed e from its strategic naming as “Janmarg” to mean “People's Way”, to the slickly designed

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

facilities and bus stations. The brand projects an everyman's image generating broad based stakeholder support, while also offering world-class quality. The name and logo were used widely in advertising, stationary, signboards, corridor infrastructure and facilities. For easy identification and unity, all staff wear uniforms. Signage is simple and easy to understand, and bus stations have a clean and appealing aesthetic with comfort of passengers in mind. These strategies were important in making the service appealing to a broad range of users, but also in instilling a sense of pride and identity amongst staff members (Baijal, 2011).

6.2. Delhi BRT The objectives of the Delhi BRT were always unique, driven primarily by air quality improvement rather than more traditional congestion relief goals. Though these objectives are clearly aligned, they resulted in subtly different approaches to BRT design and implementation. Firstly, it meant that pollution authorities (and not transportation agencies) initially spearheaded the BRT concept and continued to oversee transportation agencies through the planning process. One of the consequences of this was that the BRT did not comply with traditional transportation planning goals. Rather than reduce congestion by offering a ‘new’ mass transit alternative, the project aimed instead to increase public transport patronage by improving the existing bus system. The significance of this important nuance should not be underrated e and is especially pertinent in Delhi, which has tended to be somewhat unfairly judged against more traditional transportation objectives. From a technical perspective however, there were many factors beyond air pollution that made Delhi a particularly suitable candidate for bus based rapid transit. The poly-nucleated city structure, existing comprehensive road network and socioeconomic diversity of users were uniquely suited to affordable bus transport that utilized existing road infrastructure. BRT technology is low cost and can be integrated with NMT facilities to provide for the additional 40% of Delhi's citizens who walk or cycle to work (Kishore, 2009). By segregating modes and removing slower moving/stopping vehicles from traffic flows the BRT was also expected to increase lane capacity, increase safety, and reduce congestion and pollution (Tiwari, 2001). One of the key advantages however was the flexibility BRT offered in meeting changing demand in quantity, quality and location of services. Unlike light rail and metro alternatives that require high fixed infrastructure investment, bus routes can be modified quickly and at nominal cost (TRIPP, 2005b). Systems can be operationalized faster, and provide great coverage and access, extending at grade into the community (Kishore, 2009). The concept of BRT emerged initially from a proposal by the Central Pollution Control Board (CPCB) for segregated bus and bicycle lanes as a means of addressing alarming air pollution levels as well as rises in traffic accidents. In 2002 an “International Workshop on High Capacity Bus Systems” was convened to bring together international experts and Indian professionals to develop the concept (TRIPP, 2002). A high level expert ‘Committee on Sustainable Transport’ (CST), headed by the Chief Secretary of Delhi, was established to direct this and other multi-modal proposals. Recommendations of the 2002 CST report included a proposal for the implementation of BRT on five major but distinct arterial corridors within Delhi, including an initial 14.5 km pilot length from Dr. Ambedkar Nagar to Moolchand (the route that would become the current Delhi BRT). The selection of this initial pilot route was based on its strategic location, single entity ownership, ROW availability, and high level of bus and two-wheeler commuters

7

(with almost 85% of commuters along the selected corridor traveling by bus or two-wheeler) (Bhatia & Jain, 2009, p. 45; Kishore, 2009). However the corridor also contained Chirag Delhi e one of the busiest intersections in Delhi with a similarly large flow of powerful car users (Baijal, 2011). A Core Group comprised of Government of Delhi officials9 were established in 2003 to monitor the implementation of Committee findings and funds were earmarked in Delhi's Annual Plan 2003e2004 for “Development of Alternative mode of transport” (TRIPP, updated). The design team adhered to several key principles. Rather than settle on an ‘easily implementable’ corridor like in Ahmedabad, the Delhi design team opted for a more difficult and highly congested corridor, believing that if they could demonstrate the success of BRT under difficult conditions they would be in a better position to argue it's applicability and merits in other places in Delhi. Secondly, to minimize disruption it was decided to work within the existing confines of the selected corridor and avoid the politics associated with land acquisition (GNCTD, 2005). Finally, the importance of equity permeated the design. The existing right of way was redesigned to provide for a dedicated bus lane, and bicycle and pedestrian facilities with the view that road space should be equitably distributed amongst people (not vehicles). Grounded in strong economic and social principles, this decision nonetheless meant that the motorized vehicle lane width used by powerful and wealthy car users was reduced, and transferred to poorer NMT users (Prof. Tiwari, interview, September 14, 2012) (TRIPP, 2005a, p. 69). The design process in Delhi was both comprehensive and thoughtful. After preliminary design and extensive consultations with relevant officials and community members a second International Workshop was sponsored by the Transport Department of the National Capital Territory of Delhi (GNCTD) in December 2005. Entitled “Bus Rapid Transit Delhi” the workshop sought input from domestic and international stakeholders. It was supplemented with feedback from presentations made at the Alternative Technologies for Public Transport Conference and Exhibition organized by the Institute of Urban Transport (IUTP) in March 2006 (Baijal, 2011). Unlike Ahmedabad however, the design process in Delhi evolved slowly and was not strongly connected to parallel city and transportation initiatives. Though the development of the BRT both coincided and was aligned with the objectives of the 2005 NUTP10 there was little linkage between the two programs. Nor did it directly use funds available under the national JnNURM program (although it was accessed for an associated bus reform and fleet upgrading program for the city). There also seems to have been little coordination with the parallel metro initiative that proceeded largely in competition with Delhi BRT plans. Finally, the design process in Delhi was lengthy taking almost four years, and therefore failed to benefit from BRT's key advantage of rapid implementation. Public documents and interviews give no insights into why design took this long, however it is clear that it caused discontinuity and hindered the momentum of the scheme. The proposal for construction of the first Delhi BRT corridor was approved in 2006 after ten years of evaluation and design. Detailed profile design and specifications for the 14.5 km pilot corridor was completed by TRIPP with overall project management led by RITES

9 Core Group comprised of Chief Secretary (Chairperson), PR. Secretary (finance), Pr. Secretary to Chief Minister, Commissioner MCD, Secretary Planning, Chairmancum-Managing Director, Delhi Transport Corporation, Jt. Commissioner of Police (Traffic), Engineer in Chief (PWD), Pr. Secretary-cum Commissioner of Transport, OSD to Chief secretary, Sh. B. I. Singal, Consultant. 10 The NUTP promoted integrated land use and transport planning, more equitable allocation of road space with people (not vehicles) as the main focus; promoted multi-modal systems and supported investment in improvement of public transport and NMT facilities.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

8

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

(Baijal, 2011). DIMTS joined the team at this late point to implement and operate the scheme. The Environmental Pollution Control Authority (EPCA) closely monitored progress. Construction works commenced in October, 2006 but faced mounting political pressure to operationalize prior to elections. The first 5.8 km of this corridor running from Dr. Ambedkar Nagar to Moolchand along Josip Broz Tito (JBT) Marg in South Delhi was opened for operations in April 2008. This was to be extended to Delhi Gate in Phase II however this additional 8.7 km has never been operated as a dedicated BRT corridor despite completion of all necessary infrastructure (DIMTS, 2010; GNCTD & DIMTS, 2008; Kishore, 2009). In the early months of operation the Delhi BRT suffered a number of setbacks, creating heavy traffic jams that have had devastating impacts on public opinion. Numerous issues contributed. Driver unfamiliarity with the new traffic patterns on the corridor caused congestion that was compounded by critical software and hardware faults in the traffic signals. Pressure to operationalize had meant systems were untested and public education was incomplete. Though over 700 public and private sector bus drivers received safety training, these drivers were not always assigned to BRT routes. Bus lanes were also congested due to a high rate of bus breakdowns, and there was considerable confusion amongst passengers about boarding/alighting at central medians resulting in pedestrian jaywalking. Enforcement of the dedicated lane was also not adequate.11 The media, supported by a powerful and antagonistic car lobby12 quickly joined forces with resident welfare associations along the corridor, to criticize the ‘corridor of chaos’ and demand that it be removed (Hidalgo & Pai, 2009; Kishore, 2009; Nupur, 2012). Television interviews were shown with drivers stuck in traffic jams, pedestrians attempting dangerous road crossings and fatality reports (Singh, 2012). There was a sense amongst motorists that road space had been ‘stolen’ from private vehicles and given to bus users. DIMTS responded by upgrading traffic signal technology, deploying additional traffic marshals and assigning security guards at bus platforms, however much of the damage was done and criticism broadened (DIMTS, 2010; Hidalgo & Pai, 2009; Kishore, 2009). Media coverage accused designers of not adequately engaging government expertise, and misunderstanding on-the-ground realities (Ghosh, 2008). According to Overdorf (2012) the system was ‘built in the wrong place’, located in a middle class enclave where car users would be disproportionately inconvenienced for the benefit of few poor. Overdorf suggests the system was launched prematurely and highlighted several missed opportunities: for example, bus frequency was not increased to capitalize on the dedicated lane, the bus service provided no link to metro, and policies did not promote bus ridership nor discourage car use (e.g. parking levies and car licensing fees) (Overdorf, 2012). Various evaluations of the Delhi BRT system suggest that much of the initial media onslaught of the system was unjust and ‘confused fact with fury’ (Ecologist, 2012; Kishore, 2009).13

11 There were also accusations that disgruntled Delhi Traffic Police deliberately contributed to congestion in order to undermine its operation. Ponnaluri (2011) reports on an ‘undocumented study on signal operations’ that ‘showed that the traffic police had, based on visual observations, permitted/prevented vehicle movement without adhering to basic traffic management principles’ (Ponnaluri, 2011), p. 273. 12 In a personal interview (9/18/2012), Dunu Roy from the Hazards Center (a citizens advocacy group in Delhi) indicated that the anti-vehicle and car lobbies had an adversarial relationship that extends over 20 years. 13 “Streetsblog” reports on the disparaging post-launch media coverage of the Delhi BRT: “BRT nightmare for school kids on way home,” “kids bear the brunt of BRT mess”, “Delhi bus corridor: Fiasco continues,” “BRT corridor chaos worse than ever”. Aaron acknowledges teething problems but suggests that media bias skewed coverage (Aaron, 2008).

Proponents point out that although car users may have suffered some initial congestion, the goal of the BRT was not to serve car users, or maximize vehicular trips, but rather to maximize passenger trips along the corridor, which they had achieved. They felt that the expectations of performance improvement were unrealistic on the demonstration corridor given its limited 5.8 km length. It was pointed out that the Metro had not been subject to same scrutiny because it did not challenge road space (Lahiri, 2012). The media also failed to consider traffic flow improvements achieved by removing slower moving and stop/go bus and bicycle traffic from the left curbside lane (Prof. Tiwari, interview, January 17, 2012). Results of a joint random perception survey of BRT commuters largely vindicated the system showing that “83% of all commuters are happy with the dedicated lane system of the BRT and want that it be continued”. The majority of support came from bus users, pedestrians and cyclists. Most interestingly (and contrary to media reporting) “only 8 per cent of the motorists and two-wheeler commuters reportedly said that BRT should be scrapped and 73 percent agreed that the project can be continued” (TH, 2008a). But the damage had been done. In January 2012, a Public Interest Litigation (PIL) case was filed against the BRT, and was followed by a court ordered evaluation of the corridor by the CSIR- Central Road Research Institute (CRRI). In the meantime the dedicated bus lane was opened to motorized vehicles, effectively closing the BRT system. The ensuing debate was highly contentious and received heavy media coverage with strong opposing views from CRRI and a DIMTS/TRIPP/Greens consortium. Ultimately it was resolved in September 2012 when the High Court decided in favor of reinstating the Delhi BRT. The dedicated lane was restored to service on September 17, 2012. Legal success notwithstanding, most experts agree that while the Delhi BRT has been successful in achieving it's key objectives with respect to making public transport safer and faster and promoting NMT, the system had not achieved the same service standards as other BRTs both in India and abroad (CSE, 2009; Hidalgo & Pai, 2012). The project team cites widespread consultation with all Delhi stakeholders as part of the planning process (TRIPP, 2005a, p. 69). However the majority of these activities occurred in two distinct installments by different agents, separated by a three-year gap in communication. The first phase of communication occurred in the planning stage. This initial phase of communication effort was managed by TRIPP and involved over forty community meetings with slum dwellers and concerned residents. Feedback was also obtained from a ‘core group’ of government and transport sector officials, utility operators, and resident welfare associations along the corridor (GNCTD, 2005; Kishore, 2009). But after this initial foray, public relations efforts for the Delhi BRT appear to have lapsed for a period. From 2005 through to 2007 there is little record of community or stakeholder activities. It was during this interlude that opposition grew and then strengthened into protests as construction activities started to generate congestion. Communication efforts were restarted just prior to project launch by the newly appointed DIMTS. Given the growing resentment against the project, this second phase of communication had to focus on trouble-shooting and mitigating protests rather than true engagement (Baijal, 2011). DIMTS used in-house resources to prepare a public outreach plan that explained and described BRT's advantages, improved its image and public acceptance. Tools included BRT brochures, education and awareness raising, cycle rallies, blogs, films, flash mobs and other social media initiatives. To familiarize users with the system, over 300,000 lists of ‘do's and don’ts' were printed and distributed over time (GNCTD & DIMTS, 2008). Interactive sessions were also organized by DIMTS, including familiarization meets, open discussions, blogs, feedback lines, and presentations at schools and community associations (DIMTS, 2010).

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

But the communication strategy developed by DIMTS was in many ways too little, too late. The three-year gap between consultations had enabled anti-BRT lobbies to take hold of the debate, shifting the initial momentum and overall acceptance of BRT that had been achieved after the first phase of communication. Awareness raising and general education about the system had also been rushed in the push to operationalize (Dr. Katyal, interview, September 19, 2012). By the time DIMTS took over operations the public outcry and fury was widespread and their impressive efforts did little to mitigate the damage (Baijal, 2011). The experience provides convincing evidence for the importance of strategic and continuous sequencing of consultations in project implementation. 7. Comparing processes A comparison of the cases suggests that the Ahmedabad and Delhi BRT planning processes adopted similar content and sequence. Evaluated in this singular dimension the planning processes were largely the same, both drawing on a wealth of studies, engaging early input from design experts and involving considerable community consultation. However when other dimensions are considered e namely planning approach, and timing e it is clear that the processes are distinctly different. Delhi was long, standalone, technically focused and sequential, with an early but discontinuous communication strategy. As a result the project lacked momentum, was unable to absorb or respond effectively to changing opinion, goals were poorly understood and lacked support, and the project ultimately suffered from widespread misunderstandings regarding on-the-ground achievements. These misperceptions fueled political protests and were ultimately successful in closing down the corridor. A court ordered review of outcomes concluded that the scheme was in the public interest and re-opened the corridor, however it will take time for BRT to overcome it's unfairly tarnished brand in Delhi. In contrast, Ahmedabad was quicker, well integrated with clear objectives, more politically oriented, pragmatic and incremental. Deliberate actions were taken to link BRT development with other initiatives to build political allegiances, and the process launched incrementally to maintain momentum and allow adaptation as the process evolved. Critically, the process engaged feedback, but remained strategic without losing sight of long-term goals. The communication strategy in Ahmedabad was particularly notable and artfully managed engaging a wide variety of interactive tools and techniques. 8. Lessons learned When framed as a three dimensional activity e involving content (or steps), approach (strategy and tactics) and timing (duration and moment of action) e the importance of the planning process to outcomes in Delhi and Ahmedabad becomes clear. But what specific lessons do the cases reveal about designing a planning process? Firstly, the cases show how planning processes can be improved with explicit consideration of strategy and tactics, or planning approach. The ‘right’ approach to planning has been discussed early and often in the planning literature (Altshuler, 1965; Banfield, 1959; Beckman, 1964; Davidoff, 1965; Etzioni, 1967; Hoch, 1995, 2002; Howe & Langdon, 2002; Lindblom, 1959; Mandelbaum, 1979; Sanyal, 2005). It is unlikely a single approach could offer a universal roadmap to the planning process for all contexts and all phases of the planning cycle. However the cases do highlight the usefulness of ‘strategic’ incrementalism as a possible strategy e an iterative style that adjusts, and evolves in phases to changing conditions, acknowledges the political process, but who's overall

9

direction is dictated by clear objectives. There has been much criticism of incremental processes since Lindbloms first described this approach in his seminal article The Science of “Muddling Through” (Lindblom, 1959). But it is possible to implement incrementally, without necessarily ‘muddling’. The Ahmedabad case shows how incrementalism can be employed strategically. The roll out of the BRT in Ahmedabad was done in a continuous and strategic manner, with small planned steps used to build support for the project, test out alternative approaches, gather feedback, reevaluate and adapt. In the presence of clear objectives, there need be very little ‘muddling’ associated with an incremental and adaptive planning process. Secondly, the cases demonstrate the critical importance of timing. Planners have tended to ignore the significance of timing to planning outcomes, assuming that planning processes can start at any point in time, and extend indefinitely. There is little reference to the interaction of process with other events or activities or the significance of process length to outcomes. The cases of BRT implementation in Delhi and Ahmedabad show how two dimensions of timing e duration as well as moment of action e are important to planning processes, and impact perceptions and outcomes. The protracted planning process in Delhi contributed to the problems faced in BRT implementation by eroding the planning process of the momentum required to implement such an innovative project. At the same time it increased it's susceptibility to changing leadership and political support. It should not be surprising that a longer process will expose projects to more election cycles and administrative changes. It is important to recognize that in these ways, political will and leadership not just influence, but are influenced by the nature and timing of the planning process. Though Ahmedabad had decidedly better initial political conditions and stronger leadership for the project, the relatively short planning process was likewise both an outcome of, and influence on political leadership. It is easy to speculate how a stalled planning process in Ahmedabad might have undermined some of this initial project support and led to entirely different outcomes. The other dimension of timing concerns the ‘moment of action’ e or the timing of when the planning process occurs. Again Delhi and Ahmedabad differ on how they approached this issue. The Delhi team was more technocratic in their approach, effectively fulfilling sequential steps in the planning process and reporting results and progress back to political leaders. But to all intents and purposes the process was isolated, and occurred independently from other developments. As a result there were a number of missed opportunities to coordinate and build momentum for the project around supportive initiatives such as the Delhi 2021 Masterplan, the National Urban Transport Policy (NUTP), or the wellfinanced JnNURM program. In contrast, the process in Ahmedabad was much more opportunistic, building on both city, state and national initiatives e such as Accessible Ahmedabad, Gujarats' Year of Urban Development, NUTP and JnNURM for funding, publicity and political buy-in. This is not to suggest that the planning process must stall, waiting until the ‘time is right’ before it proceeds. It instead advocates an approach that is proactive, placing the BRT within a wider city/regional vision, to opportunistically build political support with aligned initiatives from all levels of government and community, and to coordinate the planning process accordingly. Finally, the cases demonstrate the importance of communication to the planning process. This is hardly contentious. Communication is widely acknowledged as an important tool for engaging stakeholders, incorporating user needs, improving design and building political support. But the Ahmedabad and Delhi experiences demonstrate the importance of engaging all in the

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

10

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11

communication process e not just users, but also likely opponents. A fatal flaw in the communication strategy for Delhi was the omission of the powerful car lobby from communication efforts. Though residents, hawkers, public transport users and pedestrians were consulted on the BRT proposals, there was no active engagement of car users, or attention to the impact proposals would have on this powerful group. In contrast Ahmedabad embarked on a participatory planning process that actively engaged all users of the road corridor, and attempted to identify and mitigate impacts e ensuring that benefits accrued to all. This included the provision of additional flyovers and river crossings for motorists as part of the corridor construction. Publicity campaigns and trial rides were also targeted to groups identified as potential opponents in an attempt to win them over to the BRT project. It is not always possible to mitigate all negative impacts but proactively managing and engaging resistance is advisable. Ahmedabad in particular artfully managed their communication and offers some useful techniques for engaging stakeholders e including traditional outreach programs (workshops, expert consultations), knowledge-sharing (study tours, exchanges), technology showcasing (prototypes, free trial period, escorted trips on BRT sections) and branding (uniforms, logo, publicity). Most importantly however, it provides an example of how communication can be used pervasively throughout the planning process with the team emitting a sense of transparency and willingness to engage with media and the community. The development of vital feedback loops that ensure information collected during communication efforts is incorporated back into the planning process is critical to success. 9. Concluding thoughts Though content of the planning process in Delhi and Ahmedabad were similar, they had distinctly different approaches and timing. Where Delhi was long, technically focused and sequential; Ahmedabad was shorter, more politically oriented, pragmatic and incremental e with opportunity for feedback and adjustment. The analysis shows how the different approaches and timing of the planning process both impacted perceived and actual project outcomes directly, and through it's influence on politics, institutions and design. This paper does not aim to dispute the key role of political leadership, strong institutions or design as stand alone influences on outcomes for all types of projects. But it does contest the current pre-occupation with design in the global BRT community. The implementation of ‘gold standard’ designs may lead to higher levels of performance and quality, but it is not always warranted, and is not a precursor of success. It's time to shift the focus of BRT debate from design standards to other influences on project outcomes. This discussion should include a broadened understanding of the planning process that includes all it's dimensions e content, approach and timing. This article emerged from a doctoral research project at Columbia University and benefited from contributions from many practitioners, planners and advocates engaged in the development of BRT projects in India and elsewhere. The paper would not have been possible without the cooperation of the more than 50 individuals who agreed to share their views and experiences of the BRT systems in Ahmedabad and Delhi. The opinions expressed in this paper are those of the authors and they take sole responsibility for any errors and omissions. References Aaron, B. (2008). Bus rapid transit under fire in Delhi, India, Newsblog, Streetsblog New York City.

ADB. (2010). Urban development policy of India (Part 1). In ADB urban innovations and best Practices: Vol. April 2010. Mobility of people by bus rapid transit. Manila: Asian Development Bank. Altshuler, A. (1965). The city planning process. New York: Cornell University Press. AMC. (2007). Detailed Project Report: Ahmedabad bus rapid transit system. Submitted to the Ministry of Urban Development (MoUD), India for sanction under JNNURM. Ahmedabad: Ahmedabad Municipal Corporation (AMC) Ahmedabad Urban Development Authority (AUDA) CEPT University. AMC. (2012). Accessible Ahmedabad: An initiative of Ahmedabad municipal corporation. Ahmedabad: AMC. Baijal, A. (2011). Documentation of BRT experiences: Delhi, Ahmedabad. Delhi: World Bank. Banfield, E. (1959). Ends and means in planning. International Social Science Journal, 11, 361e368. Beckman, N. (1964). The planner as a Bureaucrat. Journal of the American Institute of Planners, 30, 323e327. Bhatia, T., & Jain, M. (2009). Bus transport in Delhi working paper No. 210. Delhi: Centre for Civil Society. Bose, R., Sperling, D., Delucchi, M. A., Nesamani, K. S., Redmond, S. L. S., Schipper, L., et al. (2001). Transportation in developing Countries: Greenhouse gas scenarios for Delhi, India. Recent Work. UC Davis: Institute of Transportation studies. CEPT. (2013). Janmarg BRTS Ahmedabad: 40th month of Commercial Operations. Janmarg Monthly Operational Report. Ahmedabad: Technical Support, CoE, CEPT University. CFTE. Bus rapid transit 101. Retrieved April 26, 2012, from http://www.cfte.org. CSE. (2009). International transport expert reviews Delhi BRT. Retrieved August 14, 2012, from Centre for Science and Environment http://www.cseindia.org/node/67. Currie, G. (2005). The demand performance of bus rapid transit. Journal of Public Transportation, 8(1), 41e55. Davidoff, P. (1965). Advocacy and pluralism in planning. Journal of the American Institute of Planners, 31(4), 277e296. DIMTS. (2010). Delhi BRT system e Lessons learnt. Delhi: DIMTS. DNA. (January 14, 2010). Janmarg take A'bad to elite global club, DNA. Ecologist, Urban 2012. (June 7, 2012). Rethinking the BRT, the urban ecologist. Retrieved from http://anurbanecologist.wordpress.com/2012/06/07/rethinkingbrt-2/. Etzioni. (1967). Mixed scanning: a ‘Third’ approach to decision-making. Public Administration Review, 27(5), 385e392. Fouracre, P., & Dunkerley, C. (2003). Mass rapid transit systems for cities in the developing world. Transport Reviews, 23(3), 299e310. Ghosh, A. (April 26, 2008). No govt engineer on board, The Times of India. BRT project. Retrieved from http://articles.timesofindia.indiatimes.com/2008-04-26/delhi/ 27751858-1-brt-project-bus-rapid-transit.pwd. GNCTD, & DIMTS. (2008). Bus rapid transit in Delhi. Delhi. GNCTD. (2005). Workshop Proceedings. Paper presented at the Bus Rapid transit System Delhi, 12e13 December, 2005, Delhi. GOI. (2011). Census of India, Census Info Dashboard. Retrieved from http://www. censusindia.gov.in/2011census/censusinfodashboard/index.html. Gow, D. D., & Morss, E. R. (1988). The notorious nine: critical problems in project implementation. World Development, 16(12), 1399e1418. Harriss, J. (2006). Power matters: Essays on institutions, politics, and Society in India. New Delhi: Oxford University Press. Hensher, D. A., & Golob, T. F. (2008). Bus rapid transit systems: a comparative assessment. Transportation, 35, 501e518. Hesse, R. (2010). Bus rapid Transit: Worldwide Success of a modular toolbox. Paper presented at the European transport conference 2010. Glasgow: Scotland. Hidalgo, D., & Carrrigan, A. (2010). Lessons learned from major bus improvements in Latin America and Asia. In EMBARQ (Ed.), Modernizing Public Transportation. Washington DC: World Resources Institute. Hidalgo, D., Custodio, P., & Graftieaux, P. (2007). A critical look at major bus improvements in Latin American and Asia: case studies of Hitches, Hic-Ups and areas for improvement: Synthesis of lessons learned. DfID and World Bank Partnership, TRISP. Hidalgo, D., & Graftieaux, P. (2008). Bus rapid transit systems in Latin America and Asia: results and difficulties in 11 cities. Transportation Research Record, 2072, 77e88. Hidalgo, D., & Gutierrez, L. (2013). BRT and BHLS around the world: explosive growth, large positive impacts and many issues outstanding. Research in Transportation Economics, 39(1), 8e13. HIdalgo, D., & Pai, M. (2009). The Delhi bus corridor: A report by EMBARQ-the WRI Center for sustainable transport. Hidalgo, D., & Pai, M. (August 13, 2012). The Importance of Delhi's Bus Corridor. Retrieved from http://blogs.wsj.com/indiarealtime/2012/08/13the-importanceof-delhis-bus-corridor/?mod¼google_news_blog. Hoch, C. (1995). What do planners do in the United States? In S. Madelbaum, L. Mazza, & R. Burchell (Eds.), Explorations in planning theory (pp. 225e240) Brunswick NJ: CUPR Press. Hoch, C. (2002). Evaluating plans Pragmatically. Planning Theory, 1, 53e74. Hossain, M. (2006). The issues and realities of BRT planning initiatives in developing Asian cities. Journal of Public Transportation, 69e87. BRT Special Edition. Howe, J., & Langdon, C. (2002). Towards a reflexive planning theory. Planning Theory, 1(3), 209e225. ITDP. (2012). In ITDP (Ed.), The BRT Standard Version 1.0 (p. 46). New York: ITDP. Iuchi, K. (2012). Cases: Ahmedabad and Bogota. Seminar on integrated spatial development. Washington DC: World Bank.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043

A. Rizvi, E. Sclar / Research in Transportation Economics xxx (2014) 1e11 Kadri, M. (2010). People's Way: Urban Mobility in Ahmedabad. Retrieved January 31, 2011http://places.designobserver.com/entry.html?entry¼12918. Kishore, V. (July 2009). Bus rapid transit: the Indian experience. RITES Journal, 23.1e23.12. Kittelson and Associates, & Consultants, Hervert S. Levinson Transportation. (2007). Bus Rapid Transit Practioner's Guide Transit Cooperative Research Program. Washington DC: Transportation Research Board. Kogdenko, N. (2011). Successfulness of bus rapid transit systems in Asia: Ex-post evaluation. Groningen: University of Groningen (Masters). Retrieved fromhttp://www.ecn.nl/docs/library/report/2011/o11013.pdf. Kost, C. (2009). Janmarg, The People's Way. Sustainable Transport, (Winter), 7e9. Kubala, J. P., & Barton, S. (2003). BRT ¼ BTR. Center for the American dream of mobility and home ownership (10e2003). Lahiri, T. (May 25, 2012). Delhi Journal: The Big, Bad BRT, India RealTime. Retrieved from http://blogs.wsj.com/indiarealtime/2012/05/25/delhi-journal-the-big-bad-brt/. Levinson, H., Zimmerman, S., Clinger, J., & Gast, J. (2003). Bus rapid transit: Synthesis of case studies. Transportation Research Record, 1841, 1e11. Levinson, H., Zimmerman, S., Clinger, J., & Rutherford, S. C. (2002). Bus rapid transit: an overview. Journal of Public Transportation, 5(2), 1e30. Lindblom, C. E. (1959). The science of “Muddling through”. Public Administration Review, 19(2 (Spring 1959)), 79e88. Mandelbaum, S. J. (1979). A complete general theory of planning is impossible. Policy Sciences, 11, 59e71. Miller, M. A., & Buckley, S. M. (2001). Bus rapid transit institutional issues: the route from research to experience. Transportation Research Record, 1760, 34e41. NIUA.. (2009). Chapter 2: Ahmedabad bus rapid transit system urban transport initiatives in India: Best practices in PPP. New Delhi: National Institute of Urban Affairs. Nupur, P. (2012). In D. Roy (Ed.), Urban transportation: Rethinking sustainability (p. 60). Delhi: Hazards Centre. Overdorf, J. (2012). India: Delhi's exorbitant public transport failure e a series of mistakes doomed Delhi's “Bus Rapid Transit” corridor before it ever got rolling. Retrieved from http://www.minnpost.com/global-post/2012/07/india-delhisexorbitant-public-transport-failure. Pai, M., & Hidalgo, D. (2009). Indian bus rapid transit systems funded by the Jawaharlal Nehru National Urban Renewal Mission. Transportation Research Record, 2114, 10e18.

11

Polzin, S. E., & Baltes, M. R. (2002). Bus rapid transit: a viable alternative? Journal of Public Transportation, 5(2), 47e69. Ponnaluri, R. V. (2011). Sustainable Bus Rapid Transit Initiatives in India: the role of decisive leadership and strong institutions. Transport Policy, 18, 269e275. Sanyal, B. (2005). Planning as anticipation of resistance. Planning Theory, 4(3), 225e245. Singh, G. (2012, March 18). Bus rapid transit (BRT) in Delhi: who failed whom? Delhi Greens Blog. Retrieved from http://delhigreens.com/2012/03/18/bus-rapidtransit-brt-in-delhi-who-failed-whom/. TH. (April 28, 2008a). BRT caused a similar mess in Pune. The Hindu. TH. (May 22, 2008b). Belated support for BRT corridor from users. The Hindu. Retrieved from http://www.hindu.com/2008/05/22/stories/2008052253050 400.htm. Thite, D. (May 18, 2010). Ahmedabad BRTS successful due to strong political support. DNA. Thynell, M., Mohan, D., & Tiwari, G. (2010). Sustainable transport and the modernisation of urban transport in Delhi and Stockholm. Cities, 27, 421e429. Tiwari, G. (2001). Urban Transport Priorities. Paper presented at the Meeting the Transport Challenges in Southern Africa, South Africa. Tiwari, G., & Jain, D. (2010). Bus rapid transit projects in Indian cities: a status report. Built Environment, 36(3), 353e362. Tiwari, G., & Jain, D. (2012). Accessibility and safety indicators for all road users: case study Delhi BRT. Journal of Transport Geography, 22, 87e95. TRIPP. (2005b). High capacity bus systems. TRIPP. (updated). Background on the development of the BRT Project. IIT. Delhi. Retrieved from http://tripp.iitd.ernet.in/delhibrts/brts/hcbs/BACKGROUND on the development of the Brt Project.pdf. TRIPP. (2005a). First Delhi BRT corridor: a design Summary, Ambedkar Nagar to Delhi Gate. Delhi: Indian Institute of Technology Delhi. TRIPP. (January 20, 2002). Conference Proceedings. Paper presented at the International Workshop on High Capacity Bus Systems. Delhi: Pragati Maidan. Vuchic, V. R. (2007). Highway transit: Bus, Trolleybus, and bus rapid transit urban transit: Systems and technology. Hoboken, NJ: John Wiley & Sons. Wright, L. (2010). Bus rapid transit: a public transport renaissance. Built Environment, 36(3), 269e273. Wright, L., & Hook, W. (2007). Bus rapid transit planning guide. New York: Institute for Transport Development and Policy.

Please cite this article in press as: Rizvi, A., & Sclar, E., Implementing bus rapid transit: A tale of two Indian cities, Research in Transportation Economics (2014), http://dx.doi.org/10.1016/j.retrec.2014.09.043