Improving paratransit through digital innovation

Wenyu Jia, Edward Beukes, Justin Coetzee and Philip van Ryneveld

Shared taxis, informal minibuses and other forms of paratransit are a popular way of getting around in Sub Saharan Africa (SSA), yet governments and development partners have little data on how these services operate. Urban transport practitioners traditionally rely on manual vehicle and passenger count surveys conducted at terminals and on vehicles to gain insights into public transport travel demands and operations. Most SSA cities cannot afford that kind of data collection, which is both expensive and time-consuming. This means city governments have a limited understanding of system characteristics and performance (frequency of services, on-time reliability, safety, etc.), and can only respond with ad-hoc, retroactive and small-scale interventions, if any.

In recent years, the rapid rise of mobile technology has provided a unique opportunity for African cities to close the data gap in urban transport. Mobile phones provide an efficient, reliable and cost-effective way of capturing and digitizing data.  They also allow for rapid customization and variation in survey design. Building on this, several programs have leveraged mobile applications to collect data on urban transport in Sub Saharan Africa, mostly to create maps of paratransit and other urban transport services. These initiatives helped several cities get a more complete picture of their urban transport networks, but said little about how transit services were run.

So, how can we harness the power of mobile technology not just to map out transport services, but to improve the service itself? To answer this, the World Bank’s transport team, with the support of the Digital Development Partnership (DDP) and the Public-Private Infrastructure Advisory Facility (PPIAF), formulated a digital approach that applies mobile app-based data collection to transport operations diagnosis.

Paratransit diagnostic data

During 2021 – 2022, the team successfully conducted demonstration projects in Maseru, Lesotho and Gaborone, Botswana in collaboration with city officials and the local paratransit associations, gathering new data on the operational characteristics of their paratransit networks and offering each city a starting point for sector reform. Additionally, the team was able to demonstrate the field data collection to the local operators, effectively transfering this knoweldge to the sector. Here are some the key findings that emerged from this exercise:

  • Paratransit users. Users in Maseru and Gaborone are mostly lower-income individuals. Female riders are disproportionately impacted by lack of security. Work is the dominant trip purpose, followed by education and personal trips.


  • Governance and regulation: The public sector institutional environment is fragmented in both cities, with no clear locus of responsibility for public transport in either case. Paratransit regulatory oversight and budgeting for transport development fall under the purview of the central government in both countries. In the majority of cases, cities tend to issue permits for new services in response to operators’ requests, regardless of whether there is true demand for the additional services.
  • Operations and service quality: Both cities have the paradox of extensive minibus networks at the cost of limited operating hours, poor accessibility, and low frequency. There is no network planning, no fare regulation, and no system in place to define the roles of minibuses vs. shared taxis or determine which mode is better suited to each corirdor/market. This results in excessive service overlaps and destructive competitive behaviors (see illustration of network duplication below). In Maseru, 4+1 sedan taxis—which carry up four passengers and one driver—have overtaken the minibus market in recent years. The large number of shared taxis plying the roads mean lower passenger loads, which in turn results in lower productivity for the 4+1’s and worsening traffic congestion.

Paratransit networks

  • Fleet: The paratransit fleet is well beyond the manufacturer’s useful life, minibus vehicles averaging 18 years of age in Maseru, and 16 years in Gaborone. They are also poorly maintained and polluting, with an adverse impact on road safety and air quality.
  • Infrastructure and facilities: The infrastructure and facilities that support the system, such as roads or transit terminals, are in dire need of improvement. Neither city has developed city-wide infrastructure supportive of public transport as the preferred transport mode. The road infrastructure needs in Maseru are more pressing. Maseru’s feeder roads are largely unsurfaced. This has a significant impact on safety and operating costs due to higher rates of wear and tear.

In Maseru and Gaborone, mobile app data collection provided a low-cost, practical and efficient way to deepen our understanding of the complex paratransit operations . This work has allowed us to identify priorities for improving urban transport operations in these two cities, including: developing road infrastructure and transit facilities that are convenient, accessible, clean, safe, and secure; rethinking regulations to improve safety and comfort, better matching supply and demand; implementing data technologies such as tracking to guide transport reform; exploring capital subsidies to support fleet renewal, including the consideration of electric vehicle technology.

With this diagnosis, the governments in Lesotho and Botswana now have a solid basis of data-driven evidence to initiate meaningful conversations with the sector and its workforce on how to transform the paratransit business model and how to do this in a more collective way. This could take the form, for instance, of a business improvement project for a limited number of operators to prove the business case. The methodology and approach from the Bank team can be replicated in other cities across Africa and beyond.

The report can be downloaded here: https://openknowledge.worldbank.org/handle/10986/37301

Previously posted at :