The Hindu Newspaper Analysis

Editorial Topic : Can States Tax Mining Activities?

 GS-3 Mains Exam : Economy

Revision Notes

Context

• Supreme Court’s Decision: The Supreme Court upheld that State legislatures have the authority to tax mineral activities within their territories, independent of Parliament’s Mines and Minerals (Development and Regulation) Act, 1957 (1957 Act).

Key Case Details

• Section 9 of the 1957 Act: Requires mining leaseholders to pay royalties to the land lessor.
• Main Question: Whether royalties paid to State governments under the 1957 Act are considered a “tax.”
• India Cement Ltd vs. Tamil Nadu Government: Origin of the dispute where Tamil Nadu imposed a cess (additional tax) on royalties despite India Cement already paying royalties.
• Supreme Court Ruling (1989): States could only collect royalties, not impose taxes on mining activities, as per the 1957 Act and the Union’s overriding authority under Entry 54 of the Union List.

Royalty vs. Tax

• Royalty: Defined as “contractual consideration” for the right to extract minerals.
• Tax: Defined as an “imposition by a sovereign authority” to fund public services.
• Distinction: Royalties are paid to a lessor for exclusive mineral rights, while taxes are imposed by the state for public welfare.

States’ Authority on Taxing Mining Activities

• Entry 50 of the State List: Grants States the exclusive power to tax mineral rights, limited by any laws Parliament passes concerning mineral development.
• Entry 54 of the Union List: Gives the Centre power to regulate mining and mineral development in public interest.
• Supreme Court’s Reasoning:
  • Royalties are not classified as taxes, hence not falling under “taxes on mineral rights” in Entry 50.
  • The 1957 Act provides States revenue through royalties but does not interfere with their power to levy taxes on mineral rights.

Centre vs. State Powers

• Centre’s Argument: Entry 50 allows Parliament to impose limitations on State taxes on mineral rights via laws like the 1957 Act.
• Court’s Clarification:
  • The Centre can regulate mining development under Entry 54 but cannot impose taxes.
  • States have the exclusive authority to tax mineral rights under Entry 50.
  • States can also tax the land where mines are located under Article 246 and Entry 49 (taxes on lands and buildings) of the State List.

Implications of the Verdict

• Potential Changes: If the Centre amends the 1957 Act, it could divest States of their taxing power on mineral rights.
• States’ Authority: States can tax mineral-bearing lands under Entry 49 of List 2, with income from land yield as a tax measure.
• Judge’s Concerns (Justice Nagarathna):
  • Central legislation like the 1957 Act aims to promote mineral development, potentially undermined by State-imposed levies.
  • States taxing mineral rights could lead to unhealthy competition, increasing mineral costs and market instability.

Conclusion

• Financial Benefits for States: Retroactive application could financially benefit mineral-rich States like West Bengal, Odisha, and Jharkhand.
• State Laws: These states have enacted local laws to impose additional taxes on mining lessees, potentially boosting their revenue significantly.

 

 

 

The Hindu Newspaper Analysis

Editorial Topic : Teaching Computers to Forget

 GS-2 Mains Exam : Education

Revision Notes

Introduction

• Issue: Policymakers struggle with complex Machine Learning (ML) models that handle vast amounts of data through Large Language Models (LLMs) and deep neural networks.
• Challenge: Difficulty for data fiduciaries to “correct, complete, update and erase” sensitive data from systems.

The Antithesis of ML

• Machine Unlearning (MUL): A solution gaining interest among researchers and companies.
  • Purpose: Adding an algorithm to AI models to identify and delete false, incorrect, discriminatory, outdated, and sensitive information.
  • Problem: Continuous data processing by LLMs creates complex data lineage, making it hard to track and remove sensitive data.
  • Risks: No sandbox approach for data processing leads to data poisoning, where hackers can insert manipulated data for biased outcomes.
  • Adoption: Companies like IBM are testing models for unlearning accuracy, reduced time, and cost efficiency.

Three Approaches to MUL Implementation

1. Private Approach:
   • Responsibility: Data fiduciaries test and apply MUL algorithms across training models for efficient data deletion.
   • Benefits: Enhances AI models and preserves users’ rights with minimal government intervention.
   • Challenges: Expertise and affordability issues may discourage smaller companies from testing solutions.
2. Public Approach:
   • Government Role: Prepare statutory blueprints through soft-law or hard-law approaches to obligate data fiduciaries.
   • Intervention Likelihood: High if MUL models achieve major breakthroughs amid rising regulations.
   • Mandates: Governments could issue guidelines under Data or AI Protection Regime for MUL implementation.
   • State-Prepared Models: Governments could create MUL models as part of Digital Public Infrastructure (DPI) for uniform application, addressing affordability and expertise issues for smaller companies.
3. International Approach:
   • Nation States: Collaborate to prepare a uniform framework for domestic adoption.
   • Geopolitical Challenges: Efficacy unclear due to international frictions.
   • Standard-Setting Organizations: Bodies like the International Electrotechnical Commission could develop MUL standards applicable across jurisdictions.

Way Forward

• Current Stage: MUL is in preliminary stages.
• Requirements: Stakeholders must address technical and regulatory considerations for effective implementation in the evolving AI landscape.