The Gig Welfare Trap: What the Code Says vs What the Cess Costs

The Gig Welfare Trap:
What the Code Says vs
What the Cess Costs

7.7 crore workers on e-Shram. 1–2% cess on every aggregator transaction. The Code is written. What sits beneath it — who pays, who collects, who claims — is structurally opaque. This report models three composite scenarios to surface what the Social Security Code 2020 + state platform-worker rules actually deliver, and to whom.

0 Cr e-Shram workforce

1–2% Cess range (state)

SCC 2020 Federal framework

Three workers.
Three failure modes.
One Code.

Three composite gig-worker scenarios — grounded in publicly-reported patterns from labour journalism, the Social Security Code 2020 text, and the Karnataka + Rajasthan platform-based gig-worker rules — surface what is measurable about the public layer, and what is structurally opaque.

The pattern that emerges is not allegation. It is structure. The cess is enforceable at the transaction. The benefit is enforceable at a counter the worker may never reach.

Composite 01

Vikram, 32

Food-delivery rider · Bengaluru
Swiggy + Zomato dual-app

Cess pass-through · price softening

Composite 02

Priya, 28

Ride-hail driver · Delhi NCR
Ola + Uber dual-app, EV switch

Cost absorption · take-home erosion

Composite 03

Suresh, 45

Last-mile logistics · Tier-3 MH
E-comm onboard, no UAN linkage

Benefits unreachable · structural exclusion

Field composites

Composite 01 · Food-delivery rider

Vikram 32 years · Bengaluru · Swiggy + Zomato dual-app

Cess on paper, under the Karnataka Platform-based Gig Workers (Social Security and Welfare) Bill 2024 framework, is 1–2% on every transaction the aggregator processes — paid by the aggregator into the state Welfare Board. The lived reality, as modelled from publicly-reported aggregator unit-economics, is that the cess does not stay inside the aggregator's margin envelope. It is passed through to the customer as a delivery surcharge — or, when that breaks demand, partially absorbed by reducing the rider per-order payout.

The food-delivery take-rate sits at 22–28% on an average ₹350 order. A 2% cess applied to gross transaction value is ~₹7 per order — a single-digit-percent dilution of the aggregator's margin per order in the best case, or roughly ₹1,000–1,400 per month off Vikram's take-home in the worst case. The composite-modelled friction is not whether the cess is collected — it is who absorbs it. The Code does not specify.

Failure mode: cess pass-through · margin compression

Composite 02 · Ride-hail driver

Priya 28 years · Delhi NCR · Ola + Uber, EV switch

Drives both Ola and Uber on the same shift, fares booked through both apps. Under the SCC 2020 + state notification framework, the cess is per-aggregator, per-transaction. When Priya completes 14 rides across two platforms in one day, two separate cess contributions are deposited against her e-Shram UAN — but only if both aggregators correctly resolve her phone number to the same UAN, which the Code does not require.

The EV switch is the second cost layer. The 2024 Delhi EV policy disincentivises new ICE commercial registrations, pushing drivers toward EV leases at ₹600–800/day versus owned-ICE economics of ~₹350/day all-in. The cess does nothing about this. The Welfare Board's first published benefit list — accident cover, hospitalisation grant, maternity benefit — is unverifiable from Priya's app. There is no in-app status, no claim-ledger surface.

Failure mode: contributor matching · benefit invisibility

Composite 03 · Last-mile logistics

Suresh 45 years · Tier-3 Maharashtra · e-comm partner-fleet

Onboarded onto an e-commerce last-mile partner-fleet that the aggregator routes through a fleet-operator intermediary. The Code recognises gig workers; it is less explicit on three-party arrangements where the aggregator pays the fleet-operator and the fleet-operator pays Suresh. Three structural items from the modelled scenario surface the operational gap:

No e-Shram UAN linkage at onboarding. The fleet-operator's KYC does not require it; the aggregator's manifest does not surface it. Without UAN, cess paid against the transaction is paid — but cannot be attached to Suresh's contributor history.
Tier-3 Welfare Board offices: staffed for sectoral-board legacy workflows (building workers, unorganised, beedi). The platform-worker rules are new, the SOP is unwritten, the claim process is documented in English on a state portal Suresh does not access.
Benefit awareness pipeline is the aggregator app's notification feed. Aggregators with thin partner-fleet integration deliver fewer state-policy notifications than full-stack fleets. Suresh learns about the benefit when the contribution period ends.

Failure mode: structural exclusion · benefits unreachable

The cess is enforceable.
The benefit is not.

Collection layer

Cess. Transaction. Aggregator. State Welfare Board.

Enforceable at the API call. Visible in aggregator P&L. Audited under SCC 2020 + state-level rules.

Delivery layer

Claim. UAN match. SOP. Worker awareness.

Worker-initiated. Not surfaced in the app. State-portal documented. Unverifiable from the contributor's side.

The Code priced the cess. It did not price the claim.

What is collected is automated. What is disbursed is bureaucratic. The asymmetry holds across food-delivery, ride-hail, and last-mile logistics. It is the predictable output of a regulatory architecture that is transaction-traceable but claim-opaque.

Cess Collected vs. Benefit Delivered — Composite Signal

Vikram — Food delivery

Cess @ 2%

₹7 / order

Take-home delta

−₹1.2K / mo

Priya — Ride-hail

Cess deposited

2 apps × 14 rides

In-app claim view

Not available

Suresh — Last-mile

Transactions / mo

~600

UAN-linked cess

0 (no linkage)

Cess collection layer

Benefit delivery layer

Three composite scenarios. Three failure modes. One signal underneath: the cess is a tax-event the aggregator's pipeline can process. The benefit is a worker-event the welfare-board pipeline cannot scale to.

Six questions the State Welfare
Boards can answer publicly.

These six questions are public-record-eligible under the SCC 2020 framework and the parallel state rules. The answers, when they arrive, will be more useful than the next aggregator-cess notification.

What is the cess-collected vs benefit-disbursed ratio for the first full year of the Karnataka Platform-based Gig Workers (Social Security and Welfare) Bill — disaggregated by aggregator, sector, and benefit category?

How are gig-worker contributors matched to their e-Shram UAN at the moment of cess collection? Is UAN linkage required at aggregator onboarding under the SCC 2020 rules — and if not, what fraction of cess contributions are currently un-attributable to a specific worker?

What is the median time from a notifiable gig-worker event (injury, hospitalisation, maternity) to disbursement under the SCC 2020 framework — disaggregated by state, claim type, sector?

Are aggregators required to disclose, in their statutory filings or public sustainability reporting, their state-by-state cess-paid vs platform-fee-charged-to-worker ratio? If not, where is this disclosure obligated?

How does the SCC 2020 cess-apportionment framework treat multi-app gig workers? When the same UAN earns through Swiggy + Zomato + Rapido on the same day, is the cess collected three times — and against which Welfare Board?

Where three-party arrangements exist — aggregator → fleet-operator → gig worker — which entity is the deemed "platform" for cess-collection purposes, and how is the worker's contributor-history preserved when fleet-operators rotate?

They priced the cess.
They left the claim opaque.

The three composite scenarios in this report are not allegations. They are the predictable shape of what happens when a welfare framework is architected for collection automation on the cess side and bureaucratic discretion on the disbursement side.

The question is not whether the cess is high enough. It is whether the benefit is reachable enough.

RAOSCAFF uses AI-modelled composite scenarios and a predict-not-recommend voice to surface what India's flagship welfare frameworks actually deliver, and to whom.

Predict-not-recommend: these are AI-modelled structural patterns, not investment, legal, or financial advice.

RAOSCAFF Intelligence — new briefs publish as investigations close.

Methodology

Composite gig-worker profiles modelled from publicly-reported patterns in labour journalism (The Ken, Moneycontrol, Bloomberg India, Mint), aggregator filings (Swiggy + Zomato listed disclosures, Ola/Uber competition-commission submissions), the Social Security Code 2020 text, the Karnataka Platform-based Gig Workers (Social Security and Welfare) Bill 2024, and the Rajasthan Platform-based Gig Workers (Registration and Welfare) Act 2023.

No individuals named. No specific events alleged. The analysis layer is RAOSCAFF's; the regulatory text is public; the composite scenarios are scenario-grounded model outputs. Aggregator margin estimates are derived from publicly-disclosed listed-company filings and credible market reportage; precise rider take-home estimates are illustrative, not prescriptive.

Sources

1. Social Security Code 2020 — Sections 113–122 (Gig Workers + Platform Workers + Welfare Board mechanics), full text via PRS Legislative Research.

2. Karnataka Platform-based Gig Workers (Social Security and Welfare) Bill 2024 — Department of Labour, Government of Karnataka.

3. Rajasthan Platform-based Gig Workers (Registration and Welfare) Act 2023 — full text + Welfare Board notification.

4. e-Shram portal statistics — Ministry of Labour & Employment, 2025–26 dashboards (7.7 crore registered workers, 47 notified gig-economy sectors).

5. Swiggy + Zomato listed-company filings — FY24-25 take-rate disclosures, food-delivery segment margins, transaction volume statistics.

6. The Ken / Moneycontrol / Mint / Bloomberg — composite reportage on aggregator cess pass-through, ride-hail EV transition economics, and last-mile partner-fleet arrangements (2024-26 coverage).

The Gig Welfare Trap: What the Code Says vsWhat the Cess Costs

Three workers.Three failure modes.One Code.