By Shobhit Shukla, Soumya Jha and Ulka Bhattacharyya
Our blogpost explores an important feature of the gig economy critical for its effective functioning viz. the ratings mechanism. We analyze the ‘ratings’- related provisions in the draft Delhi Motor Vehicles Aggregator Scheme, 2022 against the backdrop of existing regulatory frameworks, while noting certain conceptual and implementation-related challenges
On July 5, 2022, the Department of Transport, Government of NCT of Delhi (“Department”) published the draft Delhi Motor Vehicle Aggregators Scheme, 2022 (“Scheme”) for licensing and regulating aggregators providing passenger transport services; other delivery aggregators providing delivery services; and last-mile delivery service-providers in the National Capital Territory (“NCT”) of Delhi.
Despite the growth of the gig economy in India over the last few years, it is only recently that terms like ‘aggregator’ and ‘gig work’ were granted legal recognition. This should be viewed in light of the NITI Aayog report (p.24) of June 2022 estimating the number of persons engaged in the gig economy at 77 lakhs, in 2020-21, constituting 1.5% of India’s total workforce, and further estimated to rise to nearly 23.5 million persons, by 2029-30.
While past efforts in the form of the Motor Vehicles (Amendment) Act, 2019 (“Amendment Act”) and the Motor Vehicle Aggregator Guidelines, 2020 (“Central Guidelines”) have been made towards streamlining and bringing uniformity to the functioning of this sector, certain challenges persist.
In this backdrop, the present piece explores an important feature of the gig economy, which is critical for its effective functioning viz. the ratings mechanism. We critically analyze, in this respect, the ‘ratings’-related provisions in the Scheme, against the backdrop of the existing regulatory framework for road transport, as indicated above, and highlight certain conceptual and implementation-related challenges.
This is especially important in the wake of a large part of India’s booming gig economy being potentially attributable to transport-aggregators. Therefore, our discussion is both timely and relevant.
B. Tracing the evolution of the regulatory framework under the Motor Vehicles Act, 1988
In India, the Motor Vehicles Act, 1988 (“Act”), consolidates laws on, inter alia, the operation of motor vehicles, licensing of drivers and regulation of road transport. The Act envisages a shared regulatory framework, reflecting the distribution of legislative powers under the Constitution between the Centre and States on matters relating to road transport.
Recognizing the growing prominence of aggregators in the country, the Amendment Act was legislated to incorporate provisions concerning aggregators. Section 2(1A) of the Amendment Act conceives an aggregator as ‘a digital intermediary or market place for a passenger to connect with a driver for the purposes of transportation’; this is in line with the recommendations of the Parliamentary Standing Committee and thereafter, a Parliamentary Select Committee, both set up to contemplate amendments to the Act.
The Amendment Act also enables States to impose conditions on the operation of aggregators. To assist States in formulating such conditions, it allows them to follow any guidelines issued by the Central Government on the subject [Section 93(1)(iii)]; the Central Guidelines are issued pursuant to such powers.
Notably, the Central Guidelines recommend imposing a mechanism for rating driver-partners by corresponding passengers [Clause 7(2)(o)]. Further, they recommend remedial training programmes for driver-partners, where their ratings fall below certain percentiles [Clause 7(2)(p)].
West Bengal appears to be the only State so far to have framed and implemented State-level guidelines, in pursuance of the Central Guidelines. It is also interesting to note that the Delhi Government has proposed the Scheme under Section 67(3) of the Act, as against Section 93 of the Act, under which Central Guidelines and the guidelines in West Bengal, have been proposed.
Irrespective, potential issues with future State-level implementation of regulatory guidance for aggregators are fairly evident. These pertain to the varying manners in which States may conceptualize aggregators, ratings, and remedial training vis-à-vis the Central Guidelines or otherwise; even between the above-mentioned two States, these concepts have been interpreted differently. These potential discrepancies will likely impact the smooth functioning of aggregators across the country, who may be bound by different compliances across States, based on varying definitions and scopes. For instance, the thresholds for remedial training of driver-partners required to be undertaken by aggregators in Delhi and West Bengal are different. Requiring to train more driver-partners in one state compared to another state (based on such thresholds), may disincentivize aggregators.
C. The limits of Aggregators’ Roles and Responsibilities under the Scheme
Aggregators globally follow a self-regulatory mechanism with ratings of services provided by workers playing a pivotal role in maintaining this model. Broadly, across aggregator-platforms, a rating above 4.7 (out of 5) is considered satisfactory, whereas anything below 4.7 attracts a warning message, following which the number of jobs allocated to the worker decreases. Ratings below 4.5 attract deactivation. This model, found more-or-less across all aggregator-platforms, is based on the theory of distributed trust, a concept central to both e-commerce and sharing economy services or platforms.
Additionally, research notes that remote interactions made possible by aggregator-platforms raise questions of establishing trust. In this, ratings (and other reputation-systems) play a key role by collating data, processing such data, and producing quantitative measures that may help assess individual trustworthiness, minimizing the need for regulatory intervention.
Thus, governmental imposition of a ratings mechanism on privately functioning aggregators, is counter-intuitive, insofar as their business model goes. By design, aggregators are, as highlighted previously, self-regulating, and typically platforms, which match demand and supply. In this context, the Scheme’s current iteration raises interesting questions regarding the functional roles and responsibilities of aggregators. It also highlights the imperative of perhaps thinking of a more balanced division of roles and responsibilities amongst the State, aggregators, and driver-partners, considering the functional nature of aggregators as intermediaries.
D. An analysis of the provisions of the Scheme
Part I: Definitional challenges
The Scheme’s definition of aggregator includes not only traditional aggregators (who may usually not own or operate vehicles themselves), but also a different class of entities that may, in fact, own, operate or manage vehicles [Clause 2.2]. Including distinct functional classes under the aegis of aggregator, implies that the mandate on all aggregators to ensure remedial training for driver-partners with low ratings, may require further specification. Not doing so may end up inadvertently adding to aggregators’ compliance burdens.
The Scheme defines "rating" as an assessment of the quality of a completed trip availed by an end-user using the aggregator’s app, or the assessment by the end-user of the quality of delivery completed by the driver-partner on behalf of the aggregator. Ratings on aggregator-platforms are therefore a form of assessment towards attaining the goal of quality control of services provided, and thereby the successful functioning of aggregator-platforms. This critical function performed by the ratings mechanism makes it imperative for the Scheme to set out ratings’ provisions in clear and specified terms.
The Scheme’s definition of ratings is potentially overbroad. Ratings may either imply an assessment of the quality of the completed trip availed by the end-user or the quality of delivery completed by the driver-partner [Clause 2.20]. Much like the definition of “aggregators” under the Scheme, the definition of “ratings” seeks to capture ratings mechanisms of distinct sets of aggregators: this may be reconsidered while finalising the Scheme, to avoid confusion.
Part II: Obligations of Aggregators under Clause 4.1.4 of the Scheme
The manner in which Clause 4.1.4 conceptualizes the potential role and responsibilities of aggregators may pose certain implementational challenges:
(i) The provision does not clearly mention the scale based on which remedial training may be imparted to driver-partners. It simply states that driver-partners “having a rating less than 3.5 over a period of one year” should undertake remedial training. It is not clear if the score of ‘3.5’ is out of 5 or indicates a percentile. Given that the ratings mechanisms of different aggregators are designed differently, a common/relative standard functionally applicable to and comparable across all aggregators may be considered.
(ii) The provision does not specify the nature or contents of remedial training or measures. This requires specification, since making a functional intermediary responsible for across-the-board remedial training, without specifying its contours, is overbroad. Such risks are amplified absent any supplementary guidance on aspects on which remedial training may need to be imparted, and any further direction on whether all driver-partners falling below the given ratings threshold are to be imparted such training, irrespective of their period of engagement with the aggregator. Notably, the Central Guidelines limit an aggregator’s obligation to impart remedial training to only such driver-partners who have been engaged by the aggregator for a minimum duration; further, an aggregator is free to determine such duration [Clause 1(16)].
(iii) The provision misses acknowledging the role of both driver-partners and consumers in influencing ratings, by potentially imposing ratings-linked responsibilities solely on aggregators.
In any service mediated by an aggregator, the human element to the service is essentially the service-provider’s (in this case, the driver-partner’s). Ratings by consumers are often subjective; in fact, research shows that ratings reflect consumer preferences, including consumer biases. Ratings may also reflect only extreme opinions where very few reviews are left. There is an absence of regulatory mechanisms to ensure consumer understanding of the underlying quality control systems followed by aggregators, thereby making representativeness and reliability of user reviews doubtful. Further, there is recognition that consumers or end-users also share a common responsibility with governments and platforms to ensure the well-being of platform-workers. Therefore, assuming that consumer-provided ratings are solely based on factors in the aggregator’s control, may be incorrect. Consequently, imposing full responsibility on aggregators to provide corrective measures and remedial training, in this situation, is unlikely to effect lasting change.
(iv) The provision does not address the issue of what happens when a driver-partner switches aggregator-platforms, or operates across multiple aggregator-platforms. This is a possibility on account of the Scheme envisaging driver-partners operating across multiple aggregator-platforms [Note to Clause 3.3]. In this respect, the World Economic Forum has emphasized the importance of proper data portability rights for platform workers, as their ratings and reviews are similar to a resume, speaking directly to their reputation and credentials. Further, the OECD observes that “Personal ratings are usually lost when switching platforms.”
Part C. Data governance and cybersecurity concerns with the ratings-related provisions
The Scheme, in its present formulation and absent a comprehensive data protection law, leaves open certain concerns relating to data governance and cybersecurity. The Scheme contemplates an extensive mechanism for sharing of personal data relating to driver-partners, between aggregators and the Department. Such data includes information on ratings and grievances, which often are key determinants of a driver-partner’s reputation and income. Thus, there is a case to be made for such personal data to be accorded a minimum standard of protection along the principles recognized by the Supreme Court in Justice K.S. Puttaswamy (Retd.) v. Union of India, in the hands of aggregators, as well as the Department.
However, while the Scheme recognizes the potential applicability of the Information Technology Act, 2000, to aggregators, it does not envisage any specific safeguards for the protection of such data relating to driver-partners, post-submission to the Department. The absence of such safeguards reflects a lack of recognition of the potential privacy-harms that may accrue to driver-partners due to any breach of personal data relating to them.
The Scheme requires aggregators to (a) store and collect data relating to driver-partners having more than 15 percent grievances in 1 month, for at least 3 months from the date of the service provided [Clause 4.1.3]; and (b) provide quarterly reports on ratings and grievances against driver-partners to the Transport Department and make the same available for inspection by the Transport Department authorities [Clause 4.1.5]. The framing of these provisions leaves significant questions open to interpretation.
First, the Scheme neither provides a format nor sets out the precise scope of the quarterly records required under Clause 4.1.5. Illustratively, in the absence of guidance, a ‘report on ratings and grievances’ may be interpreted restrictively, as a report containing the cumulative average rating assigned to each driver-partner in the quarter; it may also be read expansively, specifying the rating assigned to each driver-partner for each trip undertaken by them in the quarter.
Second, Clause 4.1.5 is unclear as to whether the requirement to provide quarterly reports applies in relation to all driver-partners on the aggregator-platform, or only to those driver-partners having ratings less than 3.5 for a period of one year, referred to in Clause 4.1.4.
Third, while the Scheme requires aggregators to preserve grievance records for a minimum period of 3 months [Clause 4.1.3], it does not stipulate any maximum period for the preservation of such records. Interestingly, on this aspect, the Scheme diverges from the Central Guidelines, which impose a limitation period of 24 months for storage of all data relating to passengers and driver-partners using an aggregator’s services [Clause 9(4)].
Cumulatively, the unclear formulation of and potential overlap between the above provisions can make compliance unpredictable and cumbersome for aggregators. Such ambiguities can incentivize aggregators to err on the side of caution and maintain extensive records on driver-partners, beyond their extant practices, for longer periods than necessary to discharge their functions, at incremental compliance costs.
The Scheme is a timely development in the attempt to create a regulatory framework elucidating the roles and responsibilities of the aggregators, driver-partners and the Department in facilitating road transport services in the NCT of Delhi. Aggregators are vital stakeholders in this ecosystem, as are driver-partners, with each essaying distinct, yet interdependent roles.
We have discussed some aspects of the Scheme which may benefit from further consideration and deliberation, insofar as developing more clarity on stakeholder rights and responsibilities is concerned. For the future, we suggest that the following considerations may also be specifically noted, while finalizing the Scheme, to ensure that all stakeholders discharge their roles and responsibilities on fair and equitable terms.
First, on balancing responsibilities: In New York an independent drivers’ guild trains deactivated drivers at a fee; this is done in collaboration with the concerned cab-aggregator. Cab-aggregators are also known to partner with upskilling platforms to facilitate driver-partner training. Recently, in Delhi, the Department, in association with cab-aggregators, has initiated training for women driver-partners, to enhance their employment prospects. These models, which involve collaborative initiatives between aggregators, the government, and driver-partners, may be considered towards conducting remedial training for driver-partners under the Scheme.
Second, on developing knowledge of ratings mechanisms: Creating appropriate mechanisms to make ratings more reliable and representative, by imparting an understanding of the model to all stakeholders concerned, in terms of purpose and impact of ratings, may be a significant step, prior to imposing remedial training of driver-partners. Ratings, in this way, would be imparted in a more conscious and objective manner.
Third, on enhancing functional portability of driver-partners: A centralized mechanism/database may be envisioned to aggregate ratings of driver-partners operating across platforms/switching platforms. This could be envisioned under the aegis of the (yet to be implemented) Code on Social Security, 2020 (which requires workers to be enrolled on the web portal contemplated under Rule 50 of the Rules thereunder) or within the framework of the existing e-Shram Portal. Additionally, an idea worth exploring is eventually creating a portable skills, client-feedback and reputation certificate, for driver-partners, recognizable across similar platforms.
 In February 2023, the Supreme Court, in Roppen Transportation Services Pvt Ltd versus Union of India & Ors., observed that “while the [Central] Guidelines have to be borne in mind, the ultimate decision is to be arrived at by the State Government while considering whether to grant a licence and in regard to the formulation of rules in pursuance of the general rule making power under Section 96.” [Para 9]. Notably, Section 96 includes the power of States to frame schemes under sub-section (3) of Section 67.