A stakeholder analysis of the road transport system in Tanzania using a STAMP control structure

Abstract

The way the road transport system is developed in a country affects safety. This study aims to identify the roles and relationships of road transport stakeholders and to explore the understanding of control and feedback mechanisms and associated gaps influencing road safety. A System-Theoretic Accident Model and Processes (STAMP) model was applied to document and interview data (n = 30). Participants emphasised the hindrance of overlapping mandates among stakeholders on the road transport system’s operations and underlined the roles of coalitions for road safety as system enablers. Further, the withdrawal of some controls by international agencies can increase system vulnerability. Most importantly, critical control and feedback gaps were shown to increase risks for safety within the road transport system. The findings underscore the complexity of the road transport system and add to the discussion on a system’s approach to road safety.

Practitioner summary: Using a STAMP methodology, we extensively studied the road transport system in Tanzania. Road transport stakeholders were identified through the review of documents, interviews were conducted, and the main findings were discussed. Control and feedback mechanisms and associated gaps were critically presented, recommendations were proposed, and policy implications were suggested.

Keywords:

• STAMP control structure
• control and feedback
• stakeholders
• road safety
• road transport system

1. Introduction

The way the road transport system is developed in a country affects safety. A system that is well structured and coordinated can help reduce road trauma (Racioppi et al. 2004; Salmon et al. 2016; Zhang et al. 2021). Road trauma is a serious public health problem in many sub-Saharan African (SSA) countries including Tanzania (WHO 2018; Bhalla et al. 2013). And with a road traffic fatality rate of 29.2, compared to the global rate of 17.4 per 100,000 (WHO 2018), Tanzania remains one of the countries with the most unsafe roads in the world.

To improve the situation, it is argued that adopting a system perspective could help identify the roles and responsibilities of stakeholders within the road transport system; understand control and feedback mechanisms influencing road safety (Salmon, McClure, and Stanton 2012), and pointing out associated gaps (Hulme, McLean, et al. 2021; Read et al. 2023). Previous research has highlighted the complexity of the road transport system in Australia while analysing control and feedback gaps related to technologies (Read et al. 2023). How complexity-theory-based models work to help improve road safety (Salmon, McClure, and Stanton 2012; Leveson 2020; Qiao, Li, and Liu 2019; Stroh 2015; Elvik 2010; Ottino 2003) inspired the current study. The relationship between road safety problems and their causes is indirect and not obvious (Johal, Schemitsch, and Bhandari 2014; Scott-Parker, Goode, and Salmon 2015; FIA Foundation 2019). In addition, the management of road safety programs in four continents has proven detrimental (Opawski 2006) and spreads uncertainty among stakeholders; it also reduces the effectiveness of existing programs (Vogel et al. 2015). Despite many SSA countries having comprehensive road safety legislation, its enforcement remains minimal (Peden et al. 2004). Understanding the dynamics between different stakeholders through system analysis is of critical importance (Vogel et al. 2015). In a study done in Tanzania, it was argued that, despite the presence of multiple information and communication technology (ICT) among stakeholders with an interest in road safety, these ICT systems still do not share data (Kimaro et al. 2016). The study pointed out limited cooperation among stakeholders due to efforts to improve road safety being fragmented.

The current study was motivated by the continued discussions about the roles and responsibilities of stakeholders to reduce road traffic crashes (RTC) in Tanzania. Previous studies in SSA have highlighted downstream determinants including road users’ behavior (Katopola, Mashili, and Hasselberg 2022) and in particular the conduct of drivers and pedestrians (Patel et al. 2016; Minja and Chiduo 2010; Damsere-Derry, Palk, and King 2016) but there have been few, if any, studies of the whole system as one. So, this study aimed to gain deep insights and knowledge of the road transport system as a whole while recognising its complexity by studying parts of the system in isolation (Ottino 2003; McIlroy et al. 2019; Leveson 2004).

Thus, any incidence of RTC is not the result of a failure of one component in a sequence of traditional steps (Leveson 2004). In particular, the susceptibility of road transport systems in SSA countries could continue to burden public health systems if systematic measures are not in place. So, despite a significant number of available studies on a systems approach to road safety worldwide, there is still only a limited number of similar studies conducted in the context of SSA.

Additionally, modern road transport systems are complex due to socio-economic development and technological advancements (Qiao, Li, and Liu 2019). The complexity results from many integrated stakeholders with different roles and responsibilities functioning invisibly in the same system. Some studies relied on the System-Theoretic Accident Model and Process (STAMP) control structure (Patriarca et al. 2022; Hulme et al. 2023; Hulme, Stanton, et al. 2021) to describe each level of the road transport system in terms of levels of control (Leveson 2004). The STAMP control structure model is based on a hypothesis that system theory is a useful way to analyse accidents, such as that of road transport (Leveson 2020; Vogel et al. 2015; Leveson 2004). So, the control and feedback loops define communications between stakeholders (Salmon et al. 2016; Hamim et al. 2021; Staton et al. 2021) and have proven to be determinants of safety when making decisions within the road transport system (Salmon et al. 2016; Read et al. 2023; Staton et al. 2021; Mock et al. 2005).

The STAMP control structure model differs from traditional incident-based models which emphasise limited notions of linear causality relationships and make it hard to incorporate non-linear relationships, including feedback (Leveson 2004). Effectively preventing RTC in complex systems requires the use of crash incident models incorporating social systems, technology, and its underlying science. Without understanding the purpose, goals, and decision criteria used to construct and operate systems, it is impossible to understand and prevent crashes (Leveson 2004). Henceforth, it is argued that a failure to understand how complex socio-technical systems function leads to ineffective road safety strategies and unsafe roads (Salmon, McClure, and Stanton 2012).

Thus, systems-based research is still in its infancy in SSA countries, particularly in Tanzania. Also, RTCs are due to the multi-functionality of several determinants operating in a socio-technical system. The determinants might vary from one country to another but there are similarities across countries with the same context and level of socio-economic development (AfDB 2013). Therefore, the aim of this study was to identify the roles and relationships of road transport stakeholders and to explore the understanding of control and feedback mechanisms and associated gaps influencing road safety in the context of SSA.

2. Methodology

2.1. Study setting

Tanzania has a population of nearly 62 million people, making it the most populous country located entirely south of the equator (National Bureau of Statistics 2022). By 2018 Tanzania had 36,258 total kilometres (km) of road comprising trunk roads (12,222 km), regional roads (23,512 km), and district roads (524 km) (TANROADS 2018). The total number of registered vehicles for 2016 was 2,163,623 (WHO 2018). These include cars and 4-wheeled light vehicles, motorised 2-and 3-wheelers, heavy trucks, buses, and others.

2.2. Study design

A qualitative approach was opted to enable the abstraction of experts’ experience and interactions within the road transport system using a three-stage approach:

1. Identifying organisations involved in road safety.

2. Conducting expert stakeholders’ in-depth interviews.

3. Developing a STAMP control structure for the Tanzanian road transport system.

2.3. Identifying organisations involved in road safety

We systematically identified organisations and their respective roles in the road transport system in Tanzania (Hamim et al. 2021). We reviewed various documents, such as road safety strategies, policies, and other regulations; websites; and academic literature (Salmon et al. 2016) to identify the organisations (Table 1). The first author (DK) drafted the list of 87 organisations, discussed it with other authors (FM, HH, and MH), and reached a common understanding of the roles of organisations in the road transport system in Tanzania. This identification took six weeks due to multiple sources of information used. Thereafter, we shared a draft list with three local independent experts with broad experience and understanding of the road transport system in Tanzania. All of them have PhDs and occupy senior positions in their current employment. Four additional organisations were then added to the list, making a total of 91 organisations.

Table 1. Primary sources used to identify organisations.

Academic literature	Industry documents	Industry websites
Leveson (2004) Salmon, Read, and Stevens (2016) Hamim et al. (2021) Staton et al. (2021)	The National Transport Policy (Ministry of Works and Transport, United Republic of Tanzania, 2003) The National Road Safety Masterplan (Tanzania National Roads Agency, 2004) The National Road Safety Policy (Ministry of Works and Transport, United Republic of Tanzania, 2009)	www.parliament.go.tz www.mowt.go.tz www.tanroads.go.tz www.tarura.go.tz www.tbs.go.tz

2.4. Expert stakeholders’ in-depth interviews

We studied the roles and contributions of the organisations and grouped them according to their contributions to the road transport system. We were guided by the question, ‘How broad are the roles of the organizations?’ to decide who should be interviewed. We ensured stakeholders from each level of the system were represented. Hence, a total of 35 conveniently sampled expert stakeholders were invited for interviews and 30 (21 males, nine females) chose to participate (Table 2). The main challenges in this phase were regular postponement of the scheduled interviews and there were five no-shows for different reasons, including having a busy schedule despite an interest in participating.

Table 2. Demographics of study participants.

Participants	Gender and age	Current employment	Expertise
International = 3 Level 1 = 5 Level 2 = 9 Level 3 = 7 Level 4 = 4 Level 5 = 2 (n = 30)	Male = 70% (21) Female = 30% (9) Mean age = 46.5 (SD = 9.2476)	International agency = 10% (3) Higher education institutions = 3.3% (1) Public sector = 43.3% (13) Private sector = 20% (6) NGOs = 13.3% (4) Unions and user associations = 10% (3)	Transport planning and policy = 10% (3) Road safety advocacy = 16.7% (5) Engineering = 13.4% (4) Human factor = 23.3% (7) Others = 36.6% (11)

2.4.1. Interview guide

We prepared an interview guide as adapted from a study conducted in Bangladesh (Hamim et al. 2021). The first author prepared a Swahili version, and later we engaged an independent Swahili language speaking expert to refine it. Thereafter, we conducted two pilot interviews and used feedback obtained to refine the interview guide. Both English and Swahili tools were used during the interviews depending on the language proficiency level of the participants.

2.4.2. Interviews

We conducted the interviews between June and September 2022 with all participants (n = 30) at the managerial level. We ensured full confidentiality of handling information, and all the recordings and transcripts were kept in secure and separate places, with only researchers having access to them. Later, we recruited an independent person to transcribe and translate the interviews. The first author revisited and validated the transcriptions because of his fluency in both languages. The participants had a mean age of 46.5 years (SD = 9.2476 years) and their average level of experience was 18 years, with the public sector being well represented (43.3%) (Table 2).

2.5. Developing a STAMP control structure

Transcripts were summarised, codes were developed, and thematic content analysis was applied to identify control and feedback mechanisms within the road transport system. Control and feedback gaps were identified by all authors based on the interviews and d review of documents (Salmon et al. 2016; Read et al. 2023). Audio recordings and detailed notes were further reviewed to refine the model. Actors were either added, removed, or combined during the refinements phase. Also, both control and feedback mechanisms were amended by the authors and based on advice from the three experts. All authors analysed the transport system considering interrelated components through control and feedback mechanisms. Thus, the STAMP control structure model drafted focused on the system’s operation as conceptualised by Leveson (Leveson 2004). The model grouped organisations into six levels: international agencies; parliament and central government; government agencies; industry and user associations, unions, and companies; operational delivery and management; local management and supervision; and operating processes and environment. Other components outside the STAMP control structure were also noted and explained. The draft STAMP control structure was shared with all authors, discussed contents until the final draft was agreed. Thereafter, we modified control and feedback mechanisms and associated gaps, positions of organisations across the levels and validated if all organisations were captured and correctly placed in the model (Patel et al. 2016). The model was later shared with the three local independent experts mentioned in Section 2.3 who reviewed and recommended an approved model. The STAMP control structure for Tanzania is made up of two groups. The first group consists of international agencies and the second group of locally established organisations ranging from level 1 to level 5 (Table 3).

Table 3. Stakeholders in the Tanzanian road transport system.

Level	Description
International agencies	These include organisations, such as international financial institutions, international and regional bodies, and international governments and NGOs. They are high-level stakeholders supporting local entities. They include the World Bank, World Health Organisation, African Development Bank, European Union, and World Road Federation. Others are international development banks, regional community bodies, such as the Southern African Development Community (SADC), international governments and international NGOs and charities and international standard bodies.
Level 1: Parliament and central government	According to the governance structure in Tanzania, the Parliament is independent of the government and is responsible for formulating all legislations. The central government on the other hand is composed of several ministries which are high-level policymaking bodies. Some of these bodies are the President’s Office, Prime Minister’s Office, Ministry of Works and Transport, Ministry of Health, and Ministry of Home Affairs.
Level 2: Government agencies, industry and user associations, unions, and companies	These stakeholders are mandated to implement national level policies, plans, and strategies. Key stakeholders in this category include the Tanzanian National Roads Agency (TANROADS), Tanzanian Rural and Urban Roads Agency (TARURA), Police, Drugs Control Commission, and Land Transport Regulatory Authority (LATRA). Others included are industry and user associations, such as the Tanzanian Bus Owners Association (TABOA), engineers, NGOs, and Trade Unions. Likewise, private companies play a considerable role in ensuring safer roads for all.
Level 3: Operational delivery and management	This level includes several stakeholders directly involved in road safety operations and management. They include the traffic police division, emergency medical departments, health facilities, ambulance services, insurance companies, and driving schools.
Level 4: Local management and supervision	These stakeholders directly supervise or influence the driving processes. They include the police, teachers, parents, road assessors, judges and magistrates, and others.
Level 5: Operating processes and environment	This level refers to the interaction of road users with vehicles and vehicle systems. In the road transport system, there are many road users interacting with their vehicles and motorcycles, and they are included in this level. Also included is the road and related infrastructure which is part of the driving environment, such as traffic lights and road markings which may dictate the decisions of drivers and other road users.

2.6. Ethical considerations

Ethical approval has been received from the Research and Publication Ethical Committee of Muhimbili University of Health and Applied Sciences (MUHAS) in Tanzania. Permission to conduct interviews was granted by the organisations involved.

3. Results

3.1. Set up of the road transport system in Tanzania

Many participants acknowledged that the legal and institutional set-up is important. Reforms could strengthen the road transport system now and in the future. They cited the need for a safe system approach and a lead agency as strong determinants of an integrated road safety management system.

“…I think the most important thing is to adopt Safe System Approach…forget about the blame game, everyone must take responsibility as much as possible. Because if you look at them from the safe system approach, you may not really find them to be that safe. So, everyone needs to take full responsibility in their own area. And as much as possible to address this problem”. (Male participant, international)

Further, following the interviews, the primary list of 91 organisations was expanded to 114 due to 23 additional organisations being mentioned by participants. Figures 1–6 depict sections of the levels of STAMP control structure for the road transport system in Tanzania indicating both control and feedback mechanisms.

Figure 1. International agencies section of the STAMP control structure.

Figure 2. Level 1 section of the STAMP control structure.

Figure 3. Level 2 section of the STAMP control structure.

Figure 4. Level 3 section of the STAMP control structure.

Figure 5. Level 4 section of the STAMP control structure.

Figure 6. STAMP control structure for the Tanzanian road transport system.

3.2. International agencies

International agencies have a high-level influence on the ecosystem of policy formulation and system operation in Tanzania. Participants reported that international agencies do not have their own road safety projects. Nevertheless, they exert control through endorsing conventions, protocols, standards, technical guidance, and providing financial assistance to the government via organisations at level 1. For example, participants stated how specific protocols in SADC and East African Community (EAC) influence the safety and operations of the local road transport system. International agencies enhance capacity building by merging local and international experience, sharing best practices, funding through grants and loans, as well as monitor progress and evaluate the performance of road safety initiatives to supported organisations.

“…has a role to support governments to implement United Nations resolutions for road safety. We also support the preparation of guidelines and policies as well as legal issues and building capacity of different groups of road users through our established mechanism”. (Female participant, international)

Nonetheless, participants expressed that the agencies might establish a set of controls to be complied with during project execution. Non-compliance with controls could jeopardise the smooth progression of supported projects. The control could be in the form of developing standard manuals, such as grievance redress mechanisms (GRM) and the designation of special units for project implementation. Moreover, international agencies receive feedback from organisations at level 1 through government reports, inquiries, proposals, and other documents. In terms of communication channels, participants reported that their relationships are mostly formal.

3.3. Level 1: Parliament and central government

Participants commented that organisations at level 1 exert controls to organisations at level 2 through developing policies, strategies, plans, frameworks, and guidelines. Other controls are in the form of formulating and changing road safety legislation, supervising the government, and discussing critical national road transport issues and approving government revenues and spending on road transport though budgetary sessions (Figure 1). Participants also reported that key organisations at level 1 work in teams when preparing national road safety guidelines and standards with all communications following established government procedures and channels. Organisations at level 1 receive feedback from organisations at level 2 through government performance reports, financial progress reports, draft legislation, proposals and registration, licencing, and crash reports.

On the other hand, participants pointed out that despite having a critical role in the policy ecosystem, most organisations at level 1 face several barriers that impair the development of road safety initiatives. Some of the critical barriers include minimal funding, limited availability of capable human resources, as well as complexity in the legal and institutional set-up.

“…we usually face budget constraints and that forces us to communicate with them (international agencies) through writing a proposal to those donors to ask them for funds so that we can implement certain projects”. (Male participant, level 1).

Figure 2 shows the composition of organisations at level 1 of the STAMP control structure.

Moreover, participants believed the presence of a lead agency is key to enable effective coordination of the efforts of stakeholders in the road transport system. Most of them were optimistic about seeing an improved road safety management system with stakeholders having clear roles.

3.4. Level 2: Government agencies, industry and user associations, unions, and insurance companies

It was stated that, despite implementing road safety programs independently, some organisations at level 2 form several national technical committees that advise and send feedback to specific organisations at level 1 on matters related to road safety policies, directives, and standards. Participants reported that organisations at level 2 exert controls to organisations at level 3 through providing directives on the implementation of strategies, policies, procedures, and action plans; research priorities; regulations; enforcement of law and order; funding and training activities; emergence response system; as well as audits and inspections (Figure 2).

“… so, we have technical committees which are formed by experts from different organizations and our role is to be the Secretariat responsible for coordinating those experts in order to make decisions as per requirements in Tanzania”. (Male participant, level 2)

Some participants explained various measures imposed by government agencies when enforcing certain regulations. In addition, organisations at level 2 receive feedback from organisations at level 3 in the form of traffic flow data, performance reporting, crash, injury and fatality statistics, and project proposals and findings from organisations at level 3. Both formal and informal channels of communication are common during the implementation of road safety programs. Additionally, participants expressed that government organisations continuously interact with other organisations at the same level through advocacy, reports, and recommendations. Moreover, many participants commented on the roles of coalitions for road safety in influencing the operation of the systems. Noted coalitions were for civil society organisations (CSOs), private sector, and members of parliament. Participants commended the way these coalitions are formed and their contributions to road safety management initiatives in the country. Some of the contributions mentioned include advocating for reforms in the legal and institutional structure of the road transport system while providing technical and financial support to road safety initiatives.

3.5. Level 3: Operational delivery and management

Participants expressed that controls from organisations at level 3 to organisations at level 4 are in the form of infrastructure planning, training, accreditation and licencing, registration, and provision of research and maintenance and inspection recommendations. They also pointed out that organisations at level 3 receive feedback from organisations at level 4 regarding infringements, crashes, fatalities and injuries, traffic flow, and trips (Figure 4). Some concrete and systematic feedback mechanisms were reported where complaints made by some organisations at level 4 provide key feedback to organisations at level 3. In their opinions, several sophisticated interactions prevail among organisations at level 3 which facilitate systematic incidences handling.

“Sometimes our units get involved in road traffic crashes. When that happens, both the car and driver are affected somehow, though our transit cars are insured and have warranty. So, if a crash happens while the car is still in our hands, we are responsible for everything. We inform our client about the crash and our information includes a police report and photos. We then ask the client to choose another vehicle according to our terms and conditions”. (Male participant, level 3)

Furthermore, participants commented on collaborations of companies and non-state actors with public organisations when implementing road safety initiatives. According to these participants, interactions could be in the form of advocacy, reports, and recommendations. However, there were intriguing comments about channels of communication with both formal and informal channels perceived to achieve the intended results. Likewise, it was noted that knowledge and expertise on road transport systems and road safety relate more to individuals rather than organisations and disappear when these individuals change position. Thus, a change of leadership was said to limit communication flow, slowing down operations and programs.

“I think you know how this works, mostly through letters if you invite them to an event. There was a time when leadership was changing frequently, so it happened for example you already engaged this person and now comes another one. You need to restart the process”. (Female participant, level 3)

3.6. Level 4: Local management and supervision

The participants raised several points related to control and feedback loops for organisations at level 4. They reported that organisations at level 4 enact controls to organisations at level 5 through instructing standard operating procedures, requirements, contractual arrangements, resources allocation, risk controls, supervision and monitoring, targets and performance measures, and project specifications (Figure 5). Further, participants explained that organisations at level 4 receive feedback from level 5 in the form of crash reports, traffic flow data, audit reports, maintenance reports, incident reports, driving performance data, and inspection reports. Some feedback, such as inspection and incident reports are formally received often through established channels, while maintenance reports are ad hoc and mainly for drivers’ awareness of vehicle road worthiness. Vehicle inspection reports are used for different purposes, such as to facilitate the acquisition of vehicle registration numbers and certificate of registrations.

“In our communications we frequently share reports and to do that, we either use meetings or sometimes emails to establish avenues for convening physical meetings”. (Female participant, level 4)

3.7. Level 5: Operating processes and environment

Participants had different viewpoints regarding control and feedback mechanisms to organisations at level 5. Being at the bottom of STAMP control structure, organisations at level 5 do not exert any controls on organisations at other levels. Thus, controls are either derived from road designs, road users, or traffic regulations. Some of the mentioned controls include the design of roads and built environment, enforcement and other penalties, route guidance, such as signs and markings, information about rules and regulations, and both verbal and non-verbal social controls). Participants also reported several feedback mechanisms from road users which included observable road behaviours, verbal and non-verbal communication, and compliance with rules and regulations.

Moreover, participants had different opinions on compliance with traffic management controls. Some had concerns about non-compliant practices, such as speeding and overtaking in restricted areas. Other participants commented on frequent power outages which limit effective controls of traffic management systems on road users, leading to increased risks among the most vulnerable users, such as pedestrians.

“We still need some more measures to protect vulnerable road users”. (Female participant, level 5)

3.8. Control and feedback loops between non-adjacent levels

In addition to control structures between vertical adjacent levels, participants reported the existence of some organisations that exert controls and obtain feedback from non-adjacent levels. Participants stated that international agencies work with organisations at level 2 and 3 and their controls are in the form of funding, capacity building, performance monitoring, and approval and supervision.

“We have our own specific requirements for ensuring the systems that we support are quite safe internally and our own requirements based on international standards”. (Male participant, international)

According to these participants, these agencies receive feedback from organisations at level 2 and 3 in the form of reports, research findings, and advocacy. Some of the participants noted that the agencies largely work with umbrella organisations or associations rather than individual organisations. Some of the mentioned umbrella organisations included non-state actors formed through a coalition of CSOs, parliamentarians, and companies.

Also, organisations at level 1 enact controls on organisations at level 3 in the form of standards and government reporting, while receiving feedback in the form of campaigns and advocacy. Likewise, organisations at level 1 receive feedback from organisations at level 5 in the form of campaigns. Moreover, organisations at level 2 enact controls on organisations at level 4 and 5 in form of guidelines, strategies, action plans, and audits; while they receive feedback in form of crash information, reports, campaigns, payment of fines, and audit results.

“…When you understand some of their challenges, you may also figure out ways in which the situation can be improved. In some cases, it is an issue of incentives, in other cases an issue of knowledge. And it could also be the licensing or the way everything is being regulated….”. (Male participant, level 2)

Additionally, organisations at level 3 exert controls on organisations at level 5 through licencing, registration, insurance pricing, and road safety education while receiving feedback in the form of crash data, audit results, driver behaviour data, and vehicle performance data.

3.9. STAMP control structure for the road transport system in Tanzania

Our results show that complex interactions exist between organisations across levels. The interactions are in the form of controls and feedback swapped between top and bottom-level organisations. It is worth noting that control and feedback mechanisms can either be formal or informal depending on the specific mandates of organisations. Figure 6 depicts both control and feedback mechanisms across the levels using arrows.

3.10. Beyond the STAMP control structure

Throughout the interviews, participants reported different ways used to promote the agenda of road safety through informal interactions in the form of lobbying and complaints. Most participants often earmarked lobbying practices and how they help in driving their programs forward. They reported that lobbying helped them to establish continued and fruitful, regular, and project-based relationships among key stakeholders at all levels. Project-based relationships exist only during project implementation and cease when projects end. Most importantly, lobbying is embedded in regular interactions between stakeholders and influences the design and implementation of road safety initiatives at almost every level in the STAMP control structure.

Additionally, participants also commented on conflicting roles and interactions existing between stakeholders in the road transport system in Tanzania. According to them, conflicting roles are mainly caused by interwoven interests among stakeholders, and this increases complexity and limits the smooth flow of information and the functionality of the system. Furthermore, some participants noted other processes that exist when users raise complaints about the road transport system. Reported complaints request immediate intervention, including fixing road crossing facilities, road humps, or bumps in areas which are prone to crashes mostly involving pedestrians. Some of these complaints are informally communicated through bodies outside of the STAMP control structure and later escalated to organisations within the control structure for implementation.

3.11. Control structure analysis—gaps in control and feedback mechanisms

The developed STAMP control structure model for Tanzania extends the analysis of control and feedback mechanisms existing in the road transport system while pointing out associated gaps. Table 4 identifies gaps in the existing STAMP control structure. The gaps are either missing controls (two) or inadequate existing controls (five). Table 5 shows example of gaps in the current feedback measures on each level in the STAMP control structure. The gaps are either feedback delays (one) or missing feedback (six).

Table 4. Examples of identified control gaps in the control structure.

Level	Gap	Type of gap	Identified by
Level 1: Parliament and Central Government	Limited financial resources to implement planned road safety programs and interventions.	Inadequate existing controls	Interviews with key informants regarding key legislation and policies guiding the mandate of key stakeholders.
Level 2: Government agencies, industry and user associations, unions, and insurance companies	Conflicting legal and institutional set-up of key organisations leading to unclear and complex road safety jurisdictions. Limited communication regarding road traffic crash investigation information	Missing controls	Interviews with key informants and comparison to aviation sector crash investigation reports, which are mandatory and generally communicated to the public.
Level 3: Operational delivery and management	Conflicting procedures for assessing, offering, and enforcing the use of driving licences, which leave room for drivers with questionable qualifications and abilities.	Inadequate existing controls	Interviews with key informants
Level 4: Local management and supervision	Inadequate emergency rescue services particularly within the road transport system. Inadequate road safety auditors and vehicle inspectors to perform road safety audits and conduct thorough vehicle inspections.	Inadequate existing controls	Interview with key informants regarding the general emergency response system, inspection, and audit.
Level 5: Operating processes and environment	Unrated and low rated roads propelling increased risks of road traffic crashes and injuries for all road users.	Inadequate existing controls	Interview with key informants

Table 5. Examples of identified feedback gaps in the control structure.

Level	Gap	Type of gap	Identified by
International organisations	Lack of planned strategy for feedback to international agencies.	Missing feedback and feedback delays	Interview with key informants regarding communication strategy among organisations.
Level 1: Parliament and Central Government	Lack of a formalised plan of feedback to facilitate easy and timely flow of information to local organisations.	Feedback delays	Interview with key informants regarding plans for feedback
Level 2: Government agencies, industry and user associations, unions, and insurance companies	Lack of data monitoring road users’ behaviours when accessing roads.	Missing feedback	Interview with key informants regarding the behaviours of road users.
Level 3: Operational delivery and management	Unavailability of traffic flow/crash data to road users which limits their ability to analyse risks and route planning.	Missing feedback	Interview with key informants regarding road traffic risks and traffic flow/offence data and statistics.
Level 4: Local management and supervision	Unavailability of crash reports and driving performance data to driver instructors to promote safe driving.	Missing feedback	Interview with key informants regarding crash reports.
Level 5: Operating processes and environment	Lack of automated data collection on driver/vehicle behaviour and near misses/crashes.	Missing feedback	Interview with key informants about automation of road transport data.

4. Discussion

This is the first study across the SSA region to apply STAMP methodology to explore mechanisms of control and feedback and associated gaps leading to risks in the road transport system. Control and feedback mechanisms could highlight upstream and downstream determinants of road safety. So, despite having high-level policy guidelines for road safety, the road safety culture in Tanzania is still mostly characterised by reactive and short-lived interventions. Our results revealed overlapping and conflicting mandates among stakeholders in the road transport system. We also found extended controls by international agencies on organisations not only at the highest level but also at lower levels. Similarly, crucial roles and contributions to road safety were found in both non-state actors and other coalitions. Moreover, our results showed increasingly interwoven formal and informal interactions among organisations within the system while highlighting road safety as a shared responsibility between countries. Also, by applying STAMP, we have managed to show control and feedback mechanisms and associated gaps that can be beneficial in other settings, such as that of LMICs in the SSA region Thus, our results expand the knowledge of systems thinking to road safety. The main findings and identified gaps are discussed in the next paragraphs.

4.1. Overlapping mandates and operations between stakeholders

The current legal, operational, and institutional frameworks could act as barriers towards synergies in the road transport system in Tanzania. This could be the case even in similar SSA countries (Salifu and Ackaah 2012). The identified control gap is related to the conflicting legal and institutional set-ups of key organisations increasing the complexity of dealing with road safety jurisdictions. While some organisations have legal mandates to coordinate road safety programs, they lack institutional instruments to implement the given mandates. Overlapping mandates and operations between these stakeholders hinder operations leading to unrealised road safety gains. Likewise, the complexity of the road transport system could also be increased by a lack of coordination among the different actors (Leveson 2004).

Additionally, participants proposed the development of a well-coordinated road transport system that is forgiving (Leveson 2004) to all road users. Participants suggested a fully funded independent entity should be established within the government that will act as a lead agency in coordinating and controlling all road safety initiatives in Tanzania. However, one of the control gaps limiting the establishment of a lead agency could be related to limited financial resources. Likewise, organisational change could also be complicated due to factors, such as the willingness, interests, and power of actors. A concrete national plan or strategy with measurable road safety targets (Johal, Schemitsch, and Bhandari 2014) and other guidelines to reflect changes happening in the road transport sector were put forward. On the other hand, the main feedback gap noted was a lack of formalised feedback plans to facilitate the easy and timely flow of information to local organisations. The feedback gap is also intensified when communications are based on personal relationships which are affected when a particular person leaves the office. This gap demonstrates the need to rethink the best approaches to developing strategies to provide the required feedback in a timely manner.

4.2. Extension of controls by international agencies to lower levels

Our results show the willingness of international agencies to support locally designed road safety initiatives despite their limited interest in providing the necessary support. Balancing the interests of international agencies with local needs remains a challenge. Contrary to findings in the UK (Staton et al. 2021) and Australia (Salmon et al. 2016), controls by international agencies extend to organisations at lower levels. This is similar to the findings of a study conducted in Bangladesh (Hamim et al. 2021). Likewise, funding has been found to be a key control and affects system operations. Thus, it could be argued that international agencies have more control and influence over lower-level organisations in low-and-middle-income countries (LMICs) than in high-income countries (HICs). This influence affects the sustainability of road safety programs.

The risk is that more controls by international agencies could lead to system vulnerability especially when the support is withdrawn. This is echoed when projects are compelled to stop before completion or are only partially implemented. Financial dependency is perceived to impair operations and sometimes affect the quality of intended road safety outputs. This is contrary to the practice in high-income countries where road transport projects are internally financed (Salmon, McClure, and Stanton 2012; Staton et al. 2021; Hughes, Anund, and Falkmer 2015). A key feedback gap noted is the lack of a clear planning strategy for feedback to international agencies leading to ad-hoc feedback and feedback delays. This could also increase the vulnerability of the road transport system in a country with limited resources.

4.3. Contributions of coalitions to road safety management

Our results show that while several initiatives are being introduced to improve road safety in Tanzania, some challenges remain unsolved. In addition, collaborations of stakeholders through strategic coalitions, such as those of CSOs, private sector organisations, and parliamentarians have been influential in developing and operating systems. The influential capacity of coalitions is mostly derived from the collective efforts as opposed to an individualistic approach. This is in line with contributions made by user and industry associations as noted in previous studies (Salmon et al. 2016; Hamim et al. 2021; Staton et al. 2021). These coalitions bring together stakeholders from different levels and are perceived by participants to be enablers and crucial for system sustainability. However, their operations are affected by control gaps, such as unclear communication strategies regarding road traffic crash investigation information and research priorities.

Moreover, results show that some stakeholders even within the coalitions act as system managers and are therefore more visible than the rest (Salmon et al. 2016). Most importantly, this STAMP control structure for Tanzania articulates the roles of stakeholders in coalitions. The coalitions attract and bring together stakeholders with similar interests while sharing their expertise in designing and implementing road safety programs. It is worth noting that the roles of coalitions in road safety have been inadequately studied. Such coalitions have not only been fundamental in supporting government-led road safety programs but also in designing and implementing their own initiatives. Some of the contributions of coalitions include the current Ten Step plan for road infrastructure, the road safety advocacy program to reform the Road Traffic Act, 1973, and other programs that have positively impacted system functionality. However, the unavailability of traffic flow/crash data to road safety coalitions could limit their ability to analyse associated risks. This is critical because actors might not receive meaningful and reliable information to support their road safety programs.

4.4. Informal interactions among stakeholders beyond the control structure

Our results highlight the use of informal means of communication during system design and implementation. It is worth noting that informal interactions contribute to system functionality despite being invisible. Moreover, these interactions magnify system complexity due to the inclusion of other stakeholders, such as the public from outside the control structure. Nevertheless, final interventions become formalised through engaging the right stakeholders as identified in the STAMP control structure to resolve raised road safety challenges.

Furthermore, the interactions increase the complexity of the system (Salmon et al. 2016; Leveson 2004; Staton et al. 2021) despite being viewed as a way of reducing bureaucracy in the system. For example, due to delays caused by a lack of formalised feedback plans, there are risks of losing crucial road safety information and knowledge. That is because road safety knowledge and expertise are sometimes attached to individuals rather than organisations and are lost once that individual leaves the organisation or influential position. Additionally, stakeholders’ interrelationships within the road transport system remain uncoordinated.

4.5. Road safety as a shared responsibility across national boundaries

Our results revealed the existence of several conventions and protocols (EAC 2009; Yesufu 2021) within both SADC and EAC influencing the road transport systems of regional blocs. The results suggest that, despite associated potential consequences, the actual impact of the growing integration of regional blocs (Staton et al. 2021) on the safety of the road transport system in Tanzania is still undetermined. Concerns about road safety are also high because of the agreements (Dolowitz and Marsh 2000; Stone 2012) on transport regulations, the free movements of people and goods as well as access to and use of land and premises across the member states (EAC 2009). Perhaps the current system is already or could be overwhelmed by the increased motorisation resulting in a greater influx of vehicles and people. The protocols could possibly burden the road transport systems due to the increased flow of vehicles and people despite relieving business operations.

This study highlights both visible and invisible controls and feedback mechanisms within the road transport system in Tanzania and across the region (Salmon et al. 2016; Staton et al. 2021). For example, a noted control gap is related to unrated and low-rated roads which pose increased risks of road traffic crashes and injuries to all road users. This creates a challenge among road transport system managers to manage crashes due to the prevalence of unsafe roads. At the same time, this is linked to missing feedback and a lack of automated data collection on associated crashes. Thus, it could be a landmark for policymakers in Tanzania and other SSA countries to begin looking at road safety as a single, complete, and inclusive structure involving all countries. We therefore argue for further analysis of these evolving trends and their effect on current and future control and feedback mechanisms at the higher levels of the control structure. This current study sets a reference point for this analysis.

4.6. Policy implications and further research

Overall, this study strengthens the idea that the presented STAMP control structure provides crucial reflections on the strengths and limitations of the road transport system in Tanzania and stipulates important suggestions for improving the most critical road safety issues. Our findings suggest that road safety prevention strategies need to address control and feedback gaps leading to risks in the road transport system and include all relevant stakeholders while demarcating their mandates to enhance system operations. Results also suggest establishing a well-coordinated lead agency supported by concrete national strategies with clear road safety performance targets and indicators. To achieve that, a comprehensive communication strategy needs to be developed and shared with the relevant stakeholders within the system. The strategy could reduce structural limitations caused by increasingly interwoven interrelationships among stakeholders. Taken together, these results suggest that this STAMP control structure model could act as a tool to guide the allocation of resources to the most influential stakeholders in the system.

Consequently, there is a need to integrate higher-level control and feedback mechanisms when collecting and analysing road transport data and developing interventions that could solve both current and future system limitations. Hence, how safe system theory is incorporated into road transport operations remains a question to be answered.

4.7. Methodological considerations

The model presented provides an extensive discussion on controls and feedback mechanisms prevailing in the Tanzanian road transport system and the associated challenges. An assumption could be that this is not fully representative of all the stakeholders in the road transport system in Tanzania. However, we aimed at deepening the understanding of the road transport system and did not include every road safety stakeholder in a single model. The presented model provides the detailed structure of the road transport system in Tanzania, thus relevant stakeholders have been included and some with similar roles have been grouped. Additionally, the model has been validated by a group of experts in the sector and represents identified stakeholders. It is fair to say that road safety is still a complex phenomenon and can be studied using different approaches. In addition, one of the strengths of this study is its triangulation of information from industry documents, academic literature, and expert stakeholder interviews to develop the STAMP control structure model for Tanzania.

The main limitation of the study was the risk of selection bias because the organisations were conveniently sampled. This could have limited the diversity of the selected organisations across all levels as presented in the control structure; however, we ensured maximum representation of organisations at all levels, and participants were internally appointed from their respective organisations. Further, all participants occupied senior positions and had broad experience in road safety. Additionally, there was a risk of social desirability bias because participants could opt to provide favourable responses about the road transport system in Tanzania. To reduce this form of bias, we guaranteed anonymity and confidentiality of the participants’ responses while using a neutral and non-suggestive approach of questioning. At the same time, the responses from participants can be considered as crucial input and as adding to the originality of the study.

5. Conclusions

This is the first study in Tanzania and across sub-Saharan Africa to apply STAMP methodology to explore the dynamics of control and feedback mechanisms while emphasising associated gaps leading to risks in the road transport system. Our study findings demonstrated that gaps, such as limited collaborations and overlapping mandates between road safety stakeholders hinder the efficiency and functionality of the system while coalitions for road safety act as system enablers. Nevertheless, the establishment of a lead agency could be a crucial step in coordinating and controlling all road safety initiatives in Tanzania. However, the proposed change could have several complications, such as conflicting interests, limited financial resources, and unclear feedback plans. The study also showed that controls by international agencies are important for system operations but could increase system vulnerability when withdrawn. Most importantly, critical control and feedback gaps were shown to increase risks for road safety within the road transport system. These findings underscore the complexity of the road transport system and add to the discussion on a system’s approach to road safety.

Author contributions

All authors were involved in the study design. DK, FM, HH, and MH: conceptualisation and methodology. DK, HH, and MH: development of data collection tools. DK: data collection. FM: supervision and assisted with logistics in the field. DK, HH, and MH: data analysis. DK, FM, HH, and MH: interpretation of the results. DK: writing—original draft preparation. DK, FM, HH, and MH: writing, reviewing, editing, and approval of the manuscript. All authors have read and agreed to the published version of the manuscript.

Acknowledgements

The authors wish to thank all participants of the study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work has been undertaken as part of doctoral research funded by the Swedish International Development Agency (SIDA).