Turnover Trouble: A single revenue model is the blind spot of the social security code

IN NOVEMBER 2025, the Ministry of Labour and Employment notified all four labour codes, marking a significant overhaul of India’s labour law architecture. Among the four, the Code on Social Security has attracted the most attention—and for good reason. For the first time, Indian law formally recognises gig and platform workers as a distinct category, extending social security benefits such as life insurance, accident cover, and health insurance to millions.

Platform workers—the cab drivers, delivery riders, and professionals in home services who power the platform economy—do not fit neatly into the traditional employer-employee mould. They are not contractually bound to a single platform. They work on their own terms, taking up gigs as and when it suits them. The rules operationalising the Code were released on May 8, 2025. There is a structural problem hiding in plain sight that could undermine the very sustainability of this social security framework. The Code proposes that platforms contribute 1-2 per cent of their turnover to a social security fund. The word ‘turnover’ sounds straightforward. To understand why, it helps to recognise that platforms broadly fall into two categories: twosided and three-sided.

Two-sided platforms connect a service provider directly with the customer. Ridehailing apps and logistics platforms are the clearest examples. A gig worker supplies the service; the customer demands it; the platform facilitates the transaction. Three-sided platforms add another layer. Here, a gig worker is the delivery or execution arm, but there is an independent supplier—a restaurant on a food delivery app or a seller on an e-commerce marketplace. The platform sits in the middle, connecting all three. Platforms account for their revenue differently, even when they fall into the same category. Consider a ride-hailing platform. A passenger books a cab from point A to point B, the fare assumed is ₹100. On a commission based model, the platform acts as an agent. It facilitates the transaction but the actual service is rendered by the driver. Of the ₹100, ₹80 goes to the driver and ₹20 is the platform’s commission. This ₹20 is what flows into the platform’s books as revenue.

Now take a logistics platform moving a parcel over the same route for the same ₹100. Here, the platform is not an agent but is the principal service provider, legally responsible for ensuring the delivery. The full ₹100 is booked as revenue, and the ₹80 paid to the delivery partner is recorded as an operating expense. The net earnings may be identical to the ride-hailing platform, but the reported turnover is five times higher. If the Code’s contribution is pegged to turnover, the logistics platform would be charged dramatically more, simply because of how its accounting is structured.

The disproportionate contribution conundrum continues with three-sided platforms. A food delivery app may earn revenue through commissions charged to restaurants, delivery fees charged to customers, or subscription fees from both. An e-commerce player may operate as a pure marketplace—listing third-party sellers and earning a small commission—or as an inventory-led model, purchasing goods outright and selling them at a margin, in which case the full sale value appears as revenue. In both cases, the gig worker’s earnings are independent of the order value. A rider is paid based on distance covered and time taken, not on whether the order was worth ₹200 or ₹2,000. Yet, under a turnover-based contribution model, the platform’s obligation to the social security fund would swing dramatically depending on its business model, even when the workforce size and work conditions are nearly identical. The Code also provides an upper cap to the platform’s contribution in the form of 5 per cent of the worker payout which may never get triggered.

Doing a rough calculation from the annual turnover reported by the largest platforms in the country show that the platforms’ contribution per worker is highly disproportionate, the gap is more than 20x. The real cost is distributed among three players—the platforms, workers’ earnings and customers’ cost of availing the services. The one with the least elasticity will bear the major chunk of the total cost. This leads us to a crosssubsidisation problem baked into the design.

A worker on one platform would effectively subsidise a worker on another platform. Two workers doing comparable work, with comparable earnings, covered under the same insurance schemes would be contributing premiums that are wildly unequal.

The problem runs deeper for platforms that operate across multiple business verticals. Platforms would contribute to the social security fund on its total turnover, including revenue from verticals that engage no gig workers at all.

Business models are also evolving faster than any static definition can keep up with. Ride-hailing platforms are quickly moving towards SaaS-based subscription models, where drivers subscribe to usage of the software platform at a flat fee and keep the entire fare, leaving the platform with zero commission or control on fares.

ONDC, the government-backed open commerce network characterised by its opensource, protocol-based approach, extends the aggregator concept beyond single-platform intermediation to a decentralised network of interoperable platforms and applications. There is an instructive parallel in formal employment. Provident Fund contributions are calculated as a percentage of the employee’s salary—not the employer’s revenues or turnover. The link is to the worker’s earnings, not the firm’s top line. This ensures consistency and fairness across employers of all sizes and business models. The same logic should apply here. Pegging platform contributions to a function of each transaction would be a far more equitable and enforceable basis than turnover. The Code on Social Security represents a genuine legislative achievement but the turnover definition, as it stands, can deter the growth of the platform economy.

The Code should consider that a single, undifferentiated notion of turnover cannot apply equitably across two-sided and threesided platforms, across agent-based and principal-based models, across marketplace and inventory players. Getting this right is not a technical footnote. It is the difference between a social security framework that is sustainable and one that is quietly gamed before it even takes off.