Compliance Framework

Four things, and most organisations have the first two.

• Identify. Know which obligations apply, which depends on entity, state, headcount, industry and the categories of worker engaged.
• Perform. Assign an owner and a deadline to each, and make the work happen.
• Evidence. Keep the record that proves it happened, in a form retrievable for the retention period.
• Detect change. Notice when an obligation changes, when the organisation crosses a threshold, or when a new state or entity brings new obligations.

The fourth is where frameworks decay. An organisation that built a correct list in 2023 and never revisited it now has a list that predates the Labour Codes coming into force on 21 November 2025, which changed the statutes underlying most of its entries.

The third is where they fail under examination. An inspection asks for the register, the challan, the return and the wage slip. An obligation that was met but cannot be evidenced is treated as one that was not.

Sorting them by what kind of action they require is more useful than sorting them by statute, because the calendar and the owner follow the action.

Every obligation the organisation carries fits one of those seven. The value of the sort is that it makes gaps visible: an organisation with a strong deposit calendar and nothing under constitute or register is a recognisable pattern, and it is usually missing a committee it was required to form years ago.

Central obligations are uniform. Almost nothing else is.

• Professional tax exists in some states and not others, with different slabs, frequencies and formats.
• Labour welfare fund applies in some states, with contribution periods that align with nothing else.
• Shops and establishments registration, hours, leave and holidays are state matters.
• Minimum wage notifications are per state, often per zone within the state, and revised on their own timetables.
• Rules under the Codes are notified state by state, so the position on a given point depends on where the establishment sits.
• Thresholds bite per establishment or per entity, so a group with several entities can have different obligations in the same city.

So the framework is obligations by state by entity, not a single national list. The practical consequence is that opening an office in a new state is a compliance event that generates a set of obligations nobody will send a reminder about, and it should trigger a review rather than a registration form.

The four Codes came into force on 21 November 2025, replacing twenty-nine central statutes. For most organisations the substance of what has to be done changed less than the citations did.

A sensible first pass through an existing framework is therefore narrow.

• Update the statutory references in the framework itself, in policies, in handbooks and in letters. An obligation cited to the Payment of Wages Act, 1936 or the Employees' Provident Funds Act, 1952 is cited to a repealed statute.
• Re-examine anything that turned on the definition of wages, because the definition now runs consistently across the Codes and adds back excluded components above one-half of total remuneration. Structures built around a low basic are the exposure here.
• Check the state rule position, since rules under the Codes are notified state by state and the position differs.
• Revisit thresholds, since some changed in the consolidation and an establishment may now be over or under one it was not before.

What does not need rebuilding is the calendar. Due dates, deposit mechanics and return formats largely carried forward, and an organisation that was meeting its deadlines in October 2025 is very likely still meeting them.

• An owner per obligation, named, rather than a department. Obligations owned by finance in general are owned by nobody in particular.
• Derivation from the payroll run. Returns and registers prepared separately from the run drift from it, and reconciliation findings follow.
• Evidence stored with the obligation, so the proof and the task are not in different systems.
• Review triggers rather than an annual review: new state, new entity, new worker category, new payment type, headcount crossing a threshold, and any amendment.
• A short exception report rather than a long compliance dashboard. What matters is what was missed, not what was met.
• Periodic end-to-end testing on a sample, following individual employees from attendance through to the return, since the joins between systems are where failures live.

None of this requires legal expertise on the team. It requires the list to be complete and the deposits to be on time, which are a records problem and a treasury problem respectively.