Regulating the On-Demand Economy: An Agenda (Working Paper)

Regulating the On-Demand Economy: An Agenda (Working Paper)

A shorter version of this working paper appeared in the Law & Our Rights section of Bangladesh’s Daily Star. The link is here: < >  In the coming weeks and months, I hope to expand the thoughts behind this article into a more substantial piece that includes other online platform enterprises.

— x —

Within a few days of Uber formally launching its operations in Dhaka, the BRTA banned its operations on the ground that it was illegally providing a taxi/private-hire service. This impasse between a regulatory authority and a market-leading start-up ‘unicorn‘, provides an opportune moment to discuss the emergence of the ‘on-demand’ economy in Bangladesh. This article will briefly discuss what is meant by the on-demand economy before elaborating on the legal problems that certain on-demand economic actors, namely ‘ride-hailing apps’ like Uber can pose to regulatory authorities. It thereby hopes to recommend certain actions that can be taken to preserve the market value of these enterprises without harming the individuals and communities that use its services.

The On-Demand Economy

The ‘on-demand’ economy refers to the “economic activity created by digital marketplaces that fulfil consumer demand via immediate access to and convenient provisioning of goods and services.” There has been a growing interest in Bangladesh in using online platform enterprises in particular, some of which are among the largest global players in the on-demand economy. Online platform enterprises seek to distinguish themselves from many other businesses involved in e-commerce as they do not sell their own goods or services but purportedly, as tech companies, help connect persons offering goods and services to those looking for them. In exchange, the platform extracts a benefit, usually in the form of a share of the earned income or a subscription. Going from TV shows (e.g. Netflix) and music (e.g. Spotify), it is now possible to get a ride in someone’s car (e.g. Uber) or a short-stay in someone’s house (e.g. Air B & B) with a few clicks of a button and a reasonable commission. In short, in a global economy low on disposable income and high on instant gratification, they monetize the growing interest in transient, immediate streaming of goods and services over actual ownership or permanent employment.

Ride-hailing, ride-sharing and carpooling platforms (e.g. Chalo, ShareSAM etc.) have now been around for a few years in Bangladesh, with Uber being the most celebrated entrant to the market. In the tech and start-up coverage of these platforms, they have generally been touted for their potential to solve specific problems: reducing traffic congestion, lowering costs of transport, freeing logistical bottlenecks and so on. Enthusiasts laud their redemptive capacity to ‘disrupt’ the conventional means of providing services, by being more agile and responsive to customer needs, offering more flexible modes of employment and nurturing a sharing culture within communities. Others have expressed concern about how these platforms will weaken the business of conventional taxicab/private-hire companies, erode the protection of workers and impact the safety of passengers. Both perspectives have merits and both camps have equal reason to be aggrieved about existing regulations, one for it being inadequate to harness the potential of emerging internet-based enterprises and the other about the lacunae in rights and protections the operations of these enterprises expose. Using the specific example of Uber will demonstrate why this is especially true in licensing and labour legislation.


The BRTA has banned Uber for providing taxi/private hire services without the requisite documentation or permission under the Taxicab Guidelines 2010, read with the Motor Vehicles Ordinance (MVO), 1983.However, that prohibition rests on the assumption that Uber is a taxicab company. In legal actions throughout the world, Uber has presented itself as an online labour brokerage dispatching ‘independent’ drivers to passengers. Unlike orthodox taxicab companies, Uber has no fleet, no central garage/depot and, in their view, employ no drivers. In some countries, it is possible to use the app to connect non-professional drivers to passengers but in many others, it is limited to those that have private hire vehicle (PHV) licenses.

In countries like Bangladesh, strict conditions are placed on the use of motor vehicles to transport private passengers for hire/reward. One option that Uber has reportedly explored is contracting with taxicab companies. However, a taxicab company would struggle to comply with both Uber’s business model and the industry’s Guidelines. While the latter requires certain distinguishing marks (Articles খ (5)-(6), চ(1)) to be used, fixed transparent fares (Articles ঙ (1)-(2)) to be charged and mobile phones not to be used while driving (Article ঝ(9)), Uber’s business model generally uses well-conditioned but anonymous cars, imposes ‘surge’ pricing to automatically raise fares during periods of high demand and penalizes drivers who do not follow predesignated routes on their smartphones.

Instead, ride-hailing platforms could look to the burgeoning number of private hire vehicles that are driven by the individuals who own them. For them ride-hailing apps present a lucrative opportunity to bolster their business. They can feasibly do so if they have, inter alia, a professional driving license (sections 2(41), (44), Form D), a contract carriage permit that may specify a designated route, a tax token (sections 51, 63(2)(i), (xii)), third party liability insurance (sections 109-110) and their vehicle is appropriately registered, meets fitness standards and has distinguishing marks. Still, these provisions, coupled with the contractual requirements of ride-hailing apps, are quite onerous and could have a dissuasive effect on using these platforms. It is also evident that the permits and licenses granted under these laws do not contemplate the use of web/mobile applications for navigation and customer complaints or GPS tracking of vehicles. In India, retrofitting earlier regulation to extend older licensing rules for ‘radio taxi’ operators to online platforms was found to be wanting. As such, this requires fresh regulation.



For this nascent industry to grow, it must come to address the problem of who is responsible for maintaining working conditions. Courts in other jurisdictions have grappled with the question: is the driver an independent contractor, using an online platform to secure client-passengers, is he a worker of an intermediary contracted with the platform or is he employed by the platform itself?

Determining ‘worker’ status is important given that the Labour Act (LA), 2006 (as amended) guarantees a raft of protections for workers that do not automatically extend to independent contractors. These include, but are not limited to, maximum working hours (sections 100-106), leave (sections 115-118), minimum wages (section 140, 145, 148-149), unionising (section 176) etc. As such, employment status determines which party bears safety net costs.

Usually, with regard to taxicab companies, under the Taxicab Guidelines, the company is obliged to only use duly-trained and qualified drivers that they have appointed (Articles ক(14)-(15)).  This provision, coupled with the definition of worker stated in MVO, 1983 (“worker means driver” in s. 2(60)) and the Fourth Schedule of the Labour Act which states that workers includes individuals employed as drivers (clauses 27, 30), indicates a presumption that a taxi driver has an worker relationship with their taxicab company.

What complicates this picture, with the insertion of a platform and its influence over the activity of drivers, is whether taxicab/private-hire-vehicle companies or even individual driver-owners are fully in control over the transport service. The question of who the employing entity is in such situations turns on the facts, rather than on contract, and will be determined by who ‘employs’ the worker and who exercises management and control over such decisions (sections 2(49)). Judgments such as Aslam, Farrar & Others v Uber B.V. et al, Case Nos. 2202550/2015 (Aslam) delivered by an English employment tribunal on 28 October 2016 indicate that only looking at the literal word of a platform’s terms and conditions is unhelpful in determining employment status. The tribunal noted the manner in which Uber twisted contractual language and created fictions to carve out any responsibilities it might have to the driver or the passenger. However, since the law seeks to identify and protect those workers in a position of subordination and dependence (Byrne Brothers (Formwork) Ltd-v-Baird & Others [2002] ICR 667, paragraph 17) the court disregarded such legal gymnastics and looked towards the real relationship between the parties. In London, it was found that Uber interviewed and recruited drivers, instructed how they may carry out their duties, asserted its discretion to accept/decline bookings for the driver as its agent, exclusively controlled key information about the passenger, penalized the cancellation of trips and the use of routes other than the one specified by them, managed performance through a rating system and reserved the power to unilaterally vary contract terms (Aslam, paragraph 92).This was sufficient to find that a contractual relationship between the driver and the passenger could not exist and that instead there was a dependant work relationship between drivers and Uber, with the former being a ‘worker’ and the latter being an employer (Aslam, paragraph 98).

Whether this will translate into a platform being considered an employing entity by Bangladeshi authorities is uncertain. While drivers seeking ‘worker’ status may similarly claim that ride-hailing apps exert a high degree of control and management, it is possible that the platform will seek to shift responsibility onto taxicab/private-hire-vehicle companies by claiming they are de facto “contracting agencies” under section 3A, LA 2006. Drivers supplied by contracting agencies are treated as workers of the latter, except during compensation claims (section 161, LA 2006).

It is worth pointing out that, in the UK, Canada, Spain and Italy, ‘dependant’/’independent’ workers represent a third employment status, between employee and self-employed, and are not extended the same protections as employees.  A roughly equivalent status in Bangladesh would be that of ‘casual’ workers, individuals who work on an “ad-hoc basis in an establishment for work of a casual nature” (section 4(4)) and are entitled to fewer rights, e.g. no pay for extended stoppage of work.  A platform could argue that their drivers fall under this classification given that they are free to log on/log off the platform but that is doubtful.

Regardless of status, certain inconsistencies with the labour legislation would remain. For instance, section 110 restricts workers being employed in more than one establishment in one day without permission which is contrary to the aim of ride-sharing, where a person may wish to offer their driving services to more than one platform/operator (‘multi-homing’) or work in another profession altogether.

Conclusions & Recommendations

The highlighted regulatory gray areas and lacunae are merely the tip of the iceberg. The operation of ride-hailing apps may raise concerns about passenger discrimination and safety, secure handling of private data (section 63, Information and Communication Technology Act, 2006) and anti-consumer allegations for variable, arbitrary pricing. Conversely, as the platforms offer similar features and products, issues regarding copyright and patent infringement might also arise. Given the uncertainty clouding the legality of ride-hailing platforms and the fact that the Competition Commission is still findings its feet, clashes between platforms or with various drivers’ associations regarding abuse of market dominance (section 16, Competition Act 2016) are still beyond the horizon.

As daunting as it may be to regulate such a fast-changing industry, going forward, a compromise should be struck between preserving laudable protections and enabling a culture of shared commuting. One option could be introducing a new transport provider category such as “on-demand transportation technology aggregator” to Bangladesh’s road transport laws as has been done in various parts of India, through advisories and guidelines.

Following a multi-stakeholder consultation, the BRTA’s transport committees could consider issuing a new form of permit for aggregators that are granted on the conditions that permit holders maintain detailed, up-to-date records on their drivers and vehicles, ensure passengers are not discriminated against or threatened and cooperate with government authorities when required. At the same time, licensed vehicles could be exempt from having distinguishing marks, colour or a separate registration category. This permit could be complemented with guidelines setting out best practices on data protection, pricing and on maintaining a ‘level playing field’ in the road transport market. With regard to the employment status of drivers, some scholars have argued in favour of introducing third, intermediate categories in jurisdictions where they are absent. For the sake of simplicity, it may be preferable to retain the existing presumption towards ‘worker’ status – if a certain amount of time is spent using the platform or an amount of income earned. The precise terms of this can be specified under further guidelines.