India’s new gig-worker law has not yet altered life on the ground: delivery workers report opaque algorithms, falling incentives and little real access to social security. Women in the gig economy face additional safety risks and an absence of maternity protections, exposing a wide implementation gap between legislation and worker experience.

Main Finding

Despite the statutory recognition of gig and platform workers under India's Code on Social Security (2020) and nominal social-security frameworks, delivery workers in Delhi-NCR continue to experience income instability, algorithmic opacity, weakened bargaining power, and limited practical access to benefits. Effective protection requires combining legislation with algorithmic transparency, enforceable grievance mechanisms, and gender-sensitive implementation.

Key Points

• Legislative context
  • The Code on Social Security (SS Code), 2020 formally recognizes gig/platform workers, mandates aggregator contributions to a Social Security Fund, and lists benefits (life/disability cover, health, maternity, pension, accident insurance).
  • India’s gig workforce estimated at 7.7 million in 2020–21, projected to reach 23.5 million by 2029–30 (NITI Aayog).
• Worker experience
  • Initial appeal: workers join for rapid entry, perceived flexibility, and early incentive schemes that offered higher earnings.
  • Algorithmic opacity: task allocation, incentive changes, and performance metrics are black boxes; workers learn changes only from declining take-home pay.
  • Declining compensation: platforms reduce per-task pay and incentives over time (often via algorithmic tuning), forcing longer hours to maintain incomes.
  • Income instability and unpredictability: dynamic pricing, variable bonuses, and opaque deactivations create persistent uncertainty.
  • Limited access to statutory benefits: low awareness, weak registration/enforcement, and platform practices undermine SS Code provisions in practice.
  • Gendered vulnerabilities: women face additional safety risks, lack of childcare/maternity support in practice, and higher exposure to harassment.
• Recommendations from the paper
  • Algorithmic transparency and worker access to decision logic or explanations.
  • Grievance redressal mechanisms and stronger platform accountability.
  • Minimum-wage guarantees / floor protections and worker-centric algorithmic design.
  • Gender-sensitive policy implementation (safety features, maternity provisions, childcare support).

Data & Methods

• Research design: mixed methods combining qualitative primary data with critical policy/document analysis.
• Primary data:
  • Semi-structured interviews and focus groups with delivery partners (principally Blinkit and similar quick-commerce platforms) in Delhi-NCR.
  • Recruitment: purposive and snowball sampling at delivery hubs and through networks.
  • Sample: workers aged ~19–35, most with >1 year on platform; interviews 30–60 minutes in Hindi; focus groups 6–8 participants.
  • Ethical steps: informed consent, anonymization, pseudonyms, recorded and transcribed.
• Secondary data:
  • SS Code (2020), government circulars, NITI Aayog reports, ministry releases, media coverage, worker organizations’ publications, academic literature.
• Analysis:
  • Thematic analysis (Braun & Clarke framework) for interviews.
  • Critical discourse analysis of policy texts to compare legislative intent vs implementation reality.
• Limitations:
  • Focused on delivery sector and one metropolitan region (Delhi-NCR) → limited generalizability.
  • Small qualitative sample captures depth, not prevalence.
  • SS Code implementation is recent; study captures an early snapshot.
  • Platform and regulator perspectives not directly interviewed; longitudinal and comparative work needed.

Implications for AI Economics

• Modeling and measurement
  • Algorithmic management is an important labor-market mechanism: platforms’ opaque algorithms effectively set wages, hours, and allocation—necessitating models that treat algorithms as endogenous policy-makers rather than neutral matching engines.
  • Opacity introduces information asymmetries and hidden shocks to earnings; empirical work should account for non-transparent rule changes (discrete algorithmic updates) as sources of idiosyncratic risk.
  • Gender heterogeneity matters: analyses of platform labor supply, welfare, and policy effects must incorporate safety, caregiving constraints, and differential access to benefits.
• Policy design and evaluation
  • Legal recognition alone (e.g., SS Code) is insufficient; economists and policymakers must model enforcement frictions, platform compliance costs, and the political economy of monitoring.
  • Regulatory experiments (e.g., mandated transparency, minimum-pay floors, enforced aggregator contributions) are natural policy levers; randomized or phased rollouts could generate causal evidence on welfare, labor supply, and platform responses.
  • Grievance redressal and explainability mechanisms can be treated as interventions—evaluate impacts on worker turnover, earnings volatility, and collective action.
• Data and empirical strategy needs
  • Access to platform-level telemetry (task allocation logs, bonus-rule histories, deactivation reasons) is crucial to identify algorithmic rule changes and quantify their labor-market effects.
  • Use of administrative or matched worker–platform data would allow estimation of algorithmic-driven wage dynamics, heterogeneity by demographics, and interactions with statutory programs.
  • Natural experiments: exploit exogenous policy changes, court rulings, platform strikes/protests, or time-stamped algorithm updates to infer causal impacts.
• Welfare and market structure
  • Platforms’ initial subsidy-then-extraction dynamics have distributional consequences; welfare analysis should weigh short-term entry benefits against long-run lock-in and decreasing returns to labor.
  • Competition in labor supply markets (large pool of workers) can amplify platforms’ monopsony power; policy (minimum wages, contribution mandates) should be evaluated for effects on employment, hours, and consumer prices.
• Practical recommendations for AI economics research
  • Integrate algorithmic transparency variables into labor models (e.g., indicator of interpretability/explainability available to workers).
  • Prioritize mixed-methods and participatory data collection to surface mechanisms hidden from platform-reported aggregates.
  • Incorporate gender-disaggregated outcomes and safety-related non-monetary costs into welfare calculations.

Short takeaway: this paper demonstrates that algorithmic governance materially shapes gig workers' economic outcomes and that legislative recognition must be coupled with enforceable transparency and accountability measures—points that should shape empirical models, data demands, and policy evaluations in AI economics.