Four codes, one country, thirty-six compliance positions

A multi-state employer reading the labour codes will find a single, coherent national framework. It will then discover that the framework does not tell it what to do. The obligations that bite — thresholds, forms, registers, timelines, procedures — live in rules made by each State, and the States are not moving together.

Why the codes alone are not enough

Labour is on the Concurrent List, which is why India had a hundred-odd labour statutes across the Centre and the States before this consolidation, and why consolidation at the centre does not by itself simplify anything at the periphery. Each of the four codes empowers the appropriate government to frame rules, and for most establishments the appropriate government is the State.

The consequence is structural rather than transitional. It does not resolve once every State has notified its rules; the rules will differ from each other permanently, as the previous State rules did.

What actually varies

• Registration and licensing. Which authority, on what form, with what supporting material, and how long the licence runs.
• Registers and records. The codes promise consolidated registers. The prescribed formats are set by rules, and an establishment in two States may maintain two versions of what is nominally the same register.
• Returns and filing calendars. Due dates and portals differ. A single national compliance calendar will be wrong in at least one State.
• Thresholds and applicability. Several obligations attach above a headcount, and States retain room to vary some of these.
• Working hours, spread-over and shift conditions, including the conditions on which women may be employed across all shifts.
• Contractor licensing thresholds, which determine whether your vendor needs a licence and therefore whether you have a principal-employer problem.

The failure mode

The way this goes wrong is rarely dramatic. An organisation appoints a national compliance owner, builds a framework against the codes and the rules of the State where the head office sits, rolls it out, and treats the remaining States as instances of the same thing.

Eighteen months later an inspection at a smaller site produces questions the framework did not anticipate, because that State prescribed a different register and a different return. The gap is not in anyone’s understanding of the codes. It is in the assumption that the codes were the operative law.

A group does not have a labour code compliance position. It has one per establishment, and they will not all be right at the same time.

How to structure it

The organising principle is that the establishment, not the entity, is the unit of compliance.

• Build an establishment register first. Every location, its State, its headcount by category including contract labour, and the thresholds it sits above or below. Most groups discover locations that were never counted — a warehouse, a site office, a branch that grew.
• Track State rule notification as a live matter, not as a one-off. A State that has not notified its rules has not gone away; it has deferred.
• Separate the common from the local. Employment contracts, appointment letters and wage structure are group-wide and should be drafted once. Registrations, registers, returns and standing orders are local and should not be templated across States.
• Give each State an owner with a name. Distributed compliance fails when it is nobody’s specific responsibility. The head-office team can own the framework; someone must own Karnataka.
• Do not let the single-return promise mislead you. Consolidated returns are real, but they consolidate filings under correctly completed registrations. If the registrations are wrong the consolidation propagates the error rather than fixing it.

The advantage nobody uses

There is a genuine upside in the consolidation, and few organisations have taken it. Because the codes replace twenty-nine statutes, a group that rebuilds its compliance architecture now — rather than porting the old one across — can retire a great deal of accumulated process that existed only because five different Acts asked for five versions of the same information.

The organisations that treat this as a migration will carry that redundancy forward for another decade. The ones that treat it as a rebuild will not.