A company choosing between PLI, ECMS, the Semiconductor Scheme, DLI, the RDI scheme, and the newer component and capital-goods schemes is routinely told they are complementary. The disbursement architectures say otherwise. Each runs on a different eligibility trigger, a different implementing agency, and a different clawback surface. What does scheme selection actually turn on for an applicant, and where does the wrong choice prove expensive two years into the cycle?
The coherent-ladder reading of India's manufacturing scheme architecture is a presentational artefact of ministry briefings, not a feature of the schemes. PLI, ECMS, the Semiconductor Programme, DLI, the RDI scheme, and the component and capital-goods schemes notified alongside them were each brought to Cabinet at a different point over the last six years, each against a different political and fiscal moment, each within a different ministry's remit, and each built with a different instrument class. They share a family resemblance at the headline level. Below the headline, the mechanics diverge sharply enough that a compliance approach calibrated for one scheme will misfire in another, and the applicant who selects on published headline benefit alone discovers two years in that the scheme's mechanics do not match the project's cash-flow profile.
The first distinction, and the one that shapes almost everything else, is disbursement architecture. PLI for electronics and most other sectors is incremental-sales linked, paying the beneficiary a percentage of the incremental turnover above a defined base year after each claim year is closed and certified. ECMS runs on a hybrid: a capital-expenditure linked component released against investment milestones, paired with a production-linked element that behaves like PLI but at a calibrated ratio; the Empowered Committee structure is what allows these two disbursement rhythms to be processed inside a single scheme. SPECS, in its active years, was a 25 percent capital subsidy disbursed against capital-expenditure certification, with no ongoing sales linkage; that design was deliberately not carried into ECMS, and applicants who expected a SPECS-style capex subsidy in ECMS have had to recalibrate their financial models. The India Semiconductor Mission is project-linked, with fiscal support of up to 50 percent of project cost released in tranches against technology and commissioning milestones; it is closer in feel to a sovereign project-finance instrument than to a production subsidy, and the India Semiconductor Mission evaluates each project on its own commercial logic rather than through a standing eligibility matrix. DLI, split into Product Design and Deployment-Linked components, pays against design milestones and end-customer deployment evidence, which is a different evidentiary world entirely. The RDI scheme, approved in 2025, operates as a long-tenor low-interest funding instrument intermediated through Anusandhan National Research Foundation-linked vehicles rather than as a direct scheme-style subsidy. A financial model built on PLI-style cash flow assumptions applied to a SPECS-style, project-linked, or RDI-style instrument misreads the working capital requirement by an order of magnitude.
The implementing agency is the second variable, and is not interchangeable. MeitY administers mobile handset PLI, IT Hardware PLI 2.0, ECMS, SPECS legacy cases, and DLI, with IFCI Limited as the standing Project Management Agency for most of these. MoHI runs the Auto and ACC Battery PLI through a separate project management arrangement. DoP administers Pharma PLI, including the bulk drugs and medical devices schemes. DPIIT carries white goods, textiles, and the default sector incentives. MoFPI runs the Food Processing PLI. Each ministry operates through its own Project Management Agency, has its own approval cascade from the PMA to the competent authority to the Empowered Committee to the Expenditure Finance Committee, and has a different institutional appetite for interpretive flexibility in the operational guidelines. The India Semiconductor Mission adds a further layer, since the Mission is a relatively recent vehicle with a project-review orientation rather than a scheme-disbursement orientation. The same claim package, adjusted for scheme mechanics, does not land identically across these agencies, and the compliance function that has calibrated itself for one ministry's PMA cycle does not transfer without rework to another.
Eligibility triggers are the third variable and the most common source of post-hoc disqualification. PLI 1.0 for mobile operated on a production-linked authorisation architecture with negotiated base years per applicant category; ECMS defines eligible components through product segment categories that the applicant's product must fall within precisely, and the fourth tranche of approvals on 30 March 2026 extended coverage into capital equipment and supply-chain items that earlier tranches had not processed in volume; SPECS operated against a component list that was not refreshed in its active years, leaving applicants who slightly missed the nomenclature outside the scheme; the India Semiconductor Mission evaluates eligibility project-by-project on technology node, fab type, and partnership configuration. An applicant whose product, process, or HS code sits outside the eligibility definition at the time of approval cannot cure the issue by subsequent reclassification. The scheme's scope is fixed at the time of the notification and at the time of the approval, and the reclassification route that operates for customs purposes does not transfer to scheme eligibility.
The merger and exit surface is where scheme selection becomes a deal-perimeter issue rather than an operational one. Beneficiary status in most schemes is non-transferable without prior approval from the implementing agency. A change of control event, including indirect changes through holding-company transactions, triggers a fresh scrutiny cycle; in several scheme notifications, it is a disqualifying event that activates clawback. Private equity and venture capital investors acquiring stakes in scheme beneficiaries routinely encounter the change-of-control clause as a transaction condition. Exit from the scheme before its notified service period, whether through restructuring, divestment, or facility closure, carries refund-with-interest liability under several instruments. A company that selected a scheme with a seven-year service period for a business line it expected to hold for five will find the scheme structuring its eventual exit.
Scheme stacking is permitted in limited circumstances but must be structured carefully. A company operating a component unit under SPECS or ECMS and a finished-goods unit under PLI cannot claim the same indigenous value in both; cost allocation, inter-unit pricing, and transfer of intermediate product become PMA-level scrutiny subjects. Applicants routinely structure separate legal entities for the component and finished-goods units to simplify attribution, which carries its own transfer pricing and income tax consequences. The single-entity approach is workable but demands far more granular accounting architecture and is more exposed at audit. A beneficiary that intends to stack schemes without establishing this structural separation at the outset typically absorbs the accounting rework and the audit exposure as a hidden cost of the incentive.
State-level alignment is the final variable and often the last one to be confirmed. Central schemes are built with the assumption that the state in which the unit is located will provide a counterpart package: land allotment at concessional rates, industrial power tariffs, stamp duty and registration waivers, SGST reimbursement, employment-linked subventions. Different states route these through different agencies and on different timelines. A central approval that lands before the state package is secured leaves the unit economics open to the state government's evolving fiscal position. Several states have in recent cycles revised or withdrawn legacy incentive offers; an applicant that banked capital-expenditure assumptions on a particular state's legacy package has had to absorb the revision. Scheme selection that does not encompass the state layer is incomplete selection, and the allocation pattern of India Semiconductor Mission projects across Gujarat, Assam, Uttar Pradesh, Odisha, Punjab, and Andhra Pradesh reflects state-level institutional capacity to deliver the counterpart package as much as it reflects central-level scheme eligibility.
The cost of a wrong scheme choice compounds quickly. Capital expenditure is sunk against a scheme the applicant does not ultimately qualify for, and by the time the disqualification or parking is known, the alternative scheme's window has closed, its operational guidelines have moved, or its PMA has completed its current batch of approvals. A project that was unit-economic under the modelled scheme becomes uneconomic without the modelled disbursement, and the capital expenditure cannot be undone. The applicant that reads disbursement architecture first, eligibility second, and published headline benefit last will make a selection that survives the second year. The applicant that reads in the opposite order will spend the second year trying to restructure into a scheme that no longer receives them.