Locked In, Priced Out? A Case for a Competition Analysis of Gig Labour
Introduction
There is no dearth of contemporary discourse surrounding the regulation of gig work in India, and rightly so. With a myriad of events stimulating increased discussion in this domain – from the notification of the new labour codes to the recent nationwide strikes by platform workers – responses to the platform economy question have largely been framed through the lens of labour protection. In this article, I seek to approach the debate from a different vantage point, analysing whether the organisation of gig work also raises concerns that are cognisable under competition law, and whether an antitrust lens can meaningfully contribute to the regulation of gig labour markets.
I argue that the manner in which gig platforms engage with the effectively oligopsonistic gig labour market, in their capacity as buyers of labour inputs, can – and does – distort contestability in the platform economy. I further contend that such concerns fall largely outside the reach of both the Competition Act, 2002, and the proposed (now withdrawn) Draft Digital Competition Bill, 2024 (‘DCB’). On this basis, I seek to demonstrate a broader gap in how competition law in India engages with buyer-side power, and to argue specifically for a broader conceptualisation of competition regulation in the platform economy.
I proceed, first, by justifying the use of a competition law lens to analyse gig platforms and by outlining the relevance of an buyer-side/input-side analysis in this context; second, by identifying the specific competition concerns that arise from the organisation of labour on gig platforms; and finally, by examining the limitations of the Competition Act and the DCB in addressing these concerns.
Justifying a Competition Lens: Gig Platforms and Labour as an “Input”
Conceptualising Platforms as “Buyers” of Labour and the Labour Market as Oligopsonistic
At the outset, it is crucial to acknowledge that gig platforms occupy a distinctive position within digital markets. Unlike conventional online marketplaces (e.g., Amazon, Flipkart), gig platforms do not merely intermediate competition between rival sellers. They also procure labour from the labour market as an input and dynamically deploy it in the provision of services to consumers. This is relevant as conventional online marketplaces typically raise competition concerns only on the seller-side – for instance, through practices such as self-preferencing. Additionally, unlike traditional entities that hire labour through contracts, gig platforms organise labour primarily through algorithmic systems that match supply and demand in real time. These are factors we must bear in mind moving forward.
Undoubtedly, availability of labour is a critical factor influencing the competitiveness of platforms, since quality, response time, geographic coverage, satisfaction, and reliability depend on access to a large and consistent pool of labour. It is further possible that the relationship and agreement between the platform and labourer and the algorithm play an important role in shaping the availability of labour both for the platform itself and – as demonstrated later – for rival platforms. For now, it suffices to say that as the availability of labour input influences the ability of a platform to enter the market and meaningfully contest, practices that unduly distort access to this pool of labour, directly or otherwise, must be scrutinised. In India, this is compounded by the limited number of entities competing, effectively making the gig labour market for each segment an oligopsony.1
A Case for Buyer-Side Competition Analysis
Before considering the potential competition implications of this buyer-side power, it must be acknowledged that seller-side concerns (i.e., consumer-facing conduct of a firm) generally continue to dominate competition law discourse worldwide. Regardless of this trend, dynamics that distort competition through the exercise of control over input markets through the exercise of buyer-side power also arguably fall squarely within the scope of competition regulation. In fact, regulators worldwide have not been completely indifferent to non-seller-side power, as evidenced by the Robinson-Patman Act in the USA, which addresses distortions arising from the exercise of purchasing power by large players to the detriment of smaller players. In principle, even in India, the Competition Act supports this analysis. The Preamble broadly seeks to “prevent…adverse effect on competition” and to “ensure freedom of trade carried on by other participants in markets”, thereby indicating a rather broad conceptualisation of the ambit of competition regulation.
Beyond the direct market contestability implications, there are also long-term economic consequences due to anti-competitive input-side practices. Even where such practices are not directly interacting with the labour market, it has been argued that their effects are likely trickle down to the root and manifest in suppressed incomes for workers. Of particular relevance is the fact that the informal labourers that are likely to be affected typically have higher consumption functions, thus, reduced wages will likely have a direct effect on aggregate demand in the long run. A broader consideration of potential economic implications of allegedly anti-competitive conduct is well-within the realm of competition law in India, which is meant to operate “keeping in view the economic development of the country” (vide Preamble of the Competition Act). These dimensions are unlikely to be captured in purely seller-side competition analysis where the conduct and analysis undertaken are more direct.
Nevertheless, it is important that the distinction between the domains of labour law and competition law be maintained. Analysing the platform economy from a competition law perspective does not entail using competition law to directly regulate labour conditions or to redistribute bargaining power (even if this may very well arise indirectly from competition regulation!). The examination should be limited to whether the conduct of gig platforms regarding labour can impair market dynamics in ways that competition regulation is meant to address. The ability of the CCI to isolate and inquire into the competition implications of multi-faceted issues has been consistently recognised (see for instance, WhatsApp LLC v CCI ¶36-37).
Bringing labour markets within the scope of competition law would be justified by the function labour performs in the platform market. As gestured towards earlier, where labour constitutes a critical input whose availability determines entry, expansion, and effective rivalry between firms, distortions in its procurement directly affect market structure and contestability. In conventional labour markets, disparities in bargaining power are mediated through external options, collective exit, or entry by rival employers. By contrast, here, it will be argued that the convergent practices identified in gig platforms operate to foreclose these mechanisms by raising new rivals’ costs of accessing labour altogether, which directly hits their ability to enter and participate in the market. This form of exclusion operates at the level of market access, and not merely at the level of bilateral bargaining between worker and platform and the specifics or conditions of a single relationship.
It is also noteworthy that a buyer-side analysis would also align with evolving understandings of competition law, which was traditionally associated narrowly with the protection of consumer welfare. However, contemporary scholars advocate for a broader conceptualisation. For instance, Dina Waked argues that antitrust regulation is “public interest law” that should not merely remedy market disruptions and focus on efficiency. Instead, it should create a “ladder of mobility” for new firms, actively dismantle entry barriers, and ensure surplus is appropriately distributed. Domestically as well, the Report of the Committee on Digital Competition Law (‘CDCL Report’) explicitly sought to broaden the objective of competition law to include considerations of contestability, fairness, and transparency. This approach is also reflected in recent decisions of the Competition Commission of India (‘CCI’) (CDCL Report, p. 36).
Therefore, the emerging consensus is that even if a firm provides a favourable experience to end consumers, its conduct may nonetheless raise competition concerns if it restricts the ability of other market participants to compete. By extension, an analysis of gig platforms as oligopsonies, and their control over labour as an input and the potential anti-competitive effects, certainly falls within the scope of the competition law, and, as has been demonstrated, the Competition Act.
Against this backdrop, in the next section, I cull out some competition problems that appear to arise on the input-side of platforms through the exercise of their buyer-power in the gig labour market.
Identifying the Competition Issue: Locked in, Priced Out
The platform economy in India spans multiple segments, most notably on-demand delivery/quick-commerce for food and groceries, ride-hailing, and personal services. As mentioned earlier, market structures are well known to be typically duopolistic or, at best, oligopolistic. As a result, few players in each segment compete to attract labour.
A review of data collated by civil society organisations, the press, and standard form Delivery Partner agreements provide some insight into the manner of allocation of gigs and attached conditions. As is well-known, gig work is algorithmically allocated. Workers are often encouraged to complete a specified number of gigs within a defined time period, with incentive structures closely tied to continued engagement on the platform. Penalties, suspension, or deactivation may be imposed for repeated rejection of gigs offered by the algorithm or for prolonged inactivity. Ratings and performance metrics have been found to further shape the allocation of future opportunities.
Therefore, there is, at the very least, a degree of self-reinforcement, i.e., more use of a platform, greater visibility to the algorithm, more access to future work. Conversely, reduced engagement results in fewer or no opportunities. Notably, this incentive structure operates independently of any formal contractual exclusivity clause and is instead embedded in the design of the platform itself.
Thus, it appears that while platform workers are formally free to multi-home or exit, the design of gig platforms imposes significant endogenous switching costs that weaken mobility in practice, such as the loss of platform-specific reputational capital, reduced algorithmic visibility, and the forfeiture of incentives. It is useful to draw a comparison with conventional labour markets, where switching costs are typically limited to the job search and retraining. However, gig platforms embed switching costs directly into the algorithm itself.
For these practices to be anti-competitive by causing ‘lock-in’ effects, it is necessary to demonstrate actual or potential softening of competition. As has been mentioned, gig platforms operate as two-sided markets in which labour participation on the input side directly conditions the ability of the platform to compete on the consumer-facing side. Due to this, in the context of ride-hailing platforms, it has been rightly argued that lock-in practices such as those described earlier raise the costs faced by potential competitors, who are unable to recruit from a labour pool that is effectively locked into incumbent platforms, thereby preventing rivals from entering the market with business models that would offer better terms (it may be beneficial to conceptualise this as analogous to the more commonly seen seller-side lock in, where the possibility of switching is effectively frustrated to shield incumbents from competitive pressure, despite the formal availability of alternatives). This is because workers are disincentivised from leaving incumbent major platforms for sustained periods, thereby concentrating effective access to labour in the hands of few players. Thus, although workers are formally free to multi-home, the aforementioned design choices substantially weaken the effectiveness of switching as a constraint on anti-competitive conduct. In fact, the harmful effects that such exploitative lock-in/mobility restraining practices have on labour competition (including in the gig economy), work conditions, and resultantly, on innovation in the broader economy, has specifically been identified and addressed by the Federal Trade Commission in the USA.
In the Indian context, the likelihood of lock-in is further compounded by two factors:
Firstly, supply of gig labour is likely rather inelastic. In low-skill or informal labour markets such as this, factors such as income dependence and switching-related friction limit the ability of workers to withdraw their services in response to deteriorating terms. This explains why such platforms may paradoxically sustain service quality despite low buyer-side competition for labour and the resultant suboptimal terms offered to workers. The reported limited impact of the recent strike by gig workers on New Year’s Eve illustrates this , perhaps evidencing that an incredibly fleeting increase in incentivisation was adequate to offset the vast majority of workers’ desire to demand sustained improvements. At most, labour supply is elastic only along the intensive margin (i.e., hours worked by workers), but inelastic along the extensive margin (i.e., market exit or entry by workers) – further evidencing a lock-in that is incapable of meaningfully disciplining conduct on a larger level. While one may argue that on New Years’ Eve, labour surplus due to incentives may have played a background role, it cannot – if at all – fully explain the widespread and rapid neutralisation of the strike, particularly given the reliance on temporary algorithmic incentives rather than any structural recruitment of new workers. This arguably indicates that platform design and income-constrained participation, rather than excess supply alone, materially weakened workers’ ability to sustain their coordinated exit by enticing workers to abandon the strike.
Secondly, findings on work allocation or incentivisation practices amongst the platforms in any given segment appears to reveal that the conduct of entities in direct competition is at the very least similar, if not identical. Therefore, the effective convergence of practices across platforms further disincentivises switching and renders it practically futile, especially when viewed in conjunction with incentive structures that reward sustained engagement with a single platform.
Thus, although switching may be an option in theory, the factors outlined above limit its ability to meaningfully influence anti-competitive platform design in practice.
Based on the foregoing analysis, there is arguably a clear case to consider competition regulation of such input-side practices of the platform economy, in relation to their interaction with the gig labour market.
Why the Competition Act and the DCB Fall Short
In this section, I seek to demonstrate that the Competition Act and the DCB both do not adequately possess the tools needed to address the issues discussed earlier. While it is acknowledged that the DCB has now been withdrawn, it would be beneficial to understand its limitations in capturing certain manifestations of anti-competitive conduct to inform what a future framework might entail.
An analysis under §3 of the Competition Act would require the existence of an “agreement” within the meaning of §2(b). §3(3) would not apply between the horizontally-related platforms as although the practices described above may appear uniform, they arise from unilateral design choices . The convergence observed here may be consistent with conscious parallelism, but conscious parallelism by itself has consistently been held insufficient to establish an agreement under §3(3). This is particularly salient in gig platform markets, where similar incentive structures and allocation mechanisms emerge as responses to common constraints and economic conditions. In fact, comparable forms of convergence and similarity in platform labour practices have been observed across jurisdictions, suggesting that such terms are not the product of coordination specific to the Indian market. Instead, they reflect broadly shared design choices inherent to the platform model itself and the most economically rational choice in response to the market.
3(4) is similarly ill-suited, as the practices in question do not necessarily stem directly from explicit clauses in the contracts between workers and platforms. While it is conceded that certain contractual provisions may provide formal backing for incentive structures or performance-linked conditions in some cases, they are neither sufficient nor determinative of the anti-competitive effects described here.
On the face of it, §4 appears to provide a more promising route, especially given increased willingness to recognise non-price factors as relevant to abuse of dominance analyses of digital markets in India (see for instance, Whatsapp LLC v CCI ¶56). However, the nature of the platform economy makes the establishment of dominance particularly difficult. Market shares in many of these sectors tend to fluctuate. Regardless, market share is only one aspect of dominance analysis under §19(4) and the consistent convergence of general practices across platforms within short timeframes, even in the recent past (see here for example), arguably suggests that market power is distributed rather than concentrated (due to an inability to “…operate independently of competitive forces prevailing in the relevant market”). The other factors outlined under §19(4) remain anchored in assumptions better suited to seller-side conduct and discrete exclusion, and do not translate very well in situations where multiple well-capitalised platforms coexist. For instance, a typical analysis of entry barriers may not be well-suited to capture the difficulty faced by a new entrant in procuring labour, as identified in Section II (which occur as a result of market-wide practices, not related to any single entity’s dominant position). Countervailing buyer power and dependence of consumers, on the other hand, retain a seller-side understanding.
There is also an increasing impulse worldwide to not merely regulate absolute/traditional dominance in the market, but relative dominance within specific market relationships. For instance, Bakhoum argues that the regulation of ‘abuse of superior bargaining position’ (in Japan) and ‘abuse of economic dependence’ (in France, Germany, and Italy) reflects the ideal of protecting the competition process and the economic freedom of all participants, not just the “minimum core” of efficiency. This largely aligns with the broader conceptualisation of competition regulation argued for at the outset as well. Therefore, it is argued that there is a strong case to regulate such lock-in effects, even if these entities are not found to be strictly dominant. The conceptual resources for such regulation do not appear to be readily available in the Competition Act.
Even if dominance were to be hypothetically established, it remains questionable whether confining scrutiny of such buyer-side conduct to ex post enforcement only after dominance is established is appropriate, particularly given the nature of effects, and evolving discourse with respect to digital markets. Apart from the peculiarities of digital markets, the issue at hand is compounded by the fact that the abuse of buyer-side power is often more embedded and more elusive than the abuse of seller-side power, making it harder to identify and correct. To be specific, the conduct in question here materialises through platform design rather than as discretely identifiable acts and manifests through a large number of small-scale individual actors, and by the time the effects become apparent and provable, market structures may already be set in stone. Thus, there is a twofold reason motivating ex ante regulation of the interaction between platforms and the gig labour market.
Recognising the limitations of ex post regulation in the broader context of digital markets, the DCB proposed ex ante obligations for Systemically Significant Digital Enterprises (‘SSDEs’), which are simply “core digital service” providers with significant market power. The DCB did include the platforms in question under core digital services (vide Schedule I). However, I argue that the obligations proposed under the DCB fall flat in relation to the gig labour market as well, given that they primarily relate to seller-side conduct, and are rather limited with respect to the kind of conduct they seek to capture. It appears that the desire to ensure ‘contestability’ appears to be understood largely only in terms of inter-firm rivalry in their capacity as sellers or providers, based on their end consumer-facing conduct. Perhaps tangentially relevant are the broadly-framed transparency obligations provided for under cl. 10, however, it remains unclear whether transparency alone would have meaningfully addressed the pervasive and complex input-side dynamics identified above.
Concluding Remarks
In conclusion, I have sought to argue that the organisation of labour in the gig economy raises competition concerns that are analytically distinct from, yet structurally intertwined with, questions of labour protection. Through a conceptualisation of platforms as buyers of labour operating within oligopsonistic gig labour markets, I have attempted to demonstrate how gig allocation and incentivisation-related practices can – and does – distort contestability in the platform economy. I have argued that such buyer-side dynamics sit uneasily within the Competition Act and are insufficiently addressed by the ex ante framework contemplated under the DCB. On a broader note, I have advocated for a broader understanding of the scope of competition regulation, particularly with regard to buyer-side distortions in input markets.
[1] Each of these geographically compete at a hyperlocal level (relying on CCI’s arguments in National Restaurant Association of India (‘NRAI’) Vs. Zomato Limited (‘Zomato’) & Others).