Twenty Years, One Amendment: The MSME Development Act and What the 2026 Bill Actually Rewrites

The law before the law

For most of independent India's history, small industry had no dedicated statute. The sector was defined under Section 11B of the Industries (Development and Regulation) Act, 1951, purely by a ceiling on investment in plant and machinery, and that ceiling was revised repeatedly as prices moved: from Rs 5 lakh in 1950 through Rs 20 lakh in 1980, Rs 60 lakh in 1991, a liberalization-era jump to Rs 3 crore in 1997, and a rollback to Rs 1 crore in 1999 that held into the 2000s, per the Ministry of Small Scale Industries' own record of the definition's history. Two structural gaps defined that era. There was no statutory category of "medium" enterprise at all, and service businesses had no size-based definition whatsoever, because the law covered only manufacturing units.

The one substantive protection that did exist was narrow. The Interest on Delayed Payments to Small Scale and Ancillary Industrial Undertakings Act, 1993, which began as a presidential ordinance in September 1992, created the first statutory compound-interest penalty for paying a small unit late. It is the direct ancestor of the delayed-payment chapter that the 2006 Act would later carry forward and strengthen.

The intellectual case for one unified law had been building since the late 1990s. The Abid Hussain Expert Committee on Small Enterprises, which reported in January 1997, argued for shifting policy away from reservation and protection toward competitiveness, and a 2000 study group led by S.P. Gupta pressed for a proper legal structure to replace a tangle of separate registrations and inspecting authorities. The resulting bill was introduced in the Lok Sabha on 12 May 2005 as the Small and Medium Enterprises Development Bill, without the word "Micro" in its title. A parliamentary standing committee examined it, and it was enacted the following year as the Micro, Small and Medium Enterprises Development Act, bringing the smallest tier formally into the statute.

What the 2006 Act built

The Micro, Small and Medium Enterprises Development Act, 2006, received presidential assent on 16 June 2006 and commenced on 2 October 2006. Its long title framed the purpose plainly: to enable the promotion, development and competitiveness of micro, small and medium enterprises. It did three durable things.

First, it created the composite legal definition the country had lacked. Section 7 classified enterprises by investment, with separate schedules for manufacturing and services: a micro manufacturing unit was capped at Rs 25 lakh of plant and machinery, a small unit at Rs 5 crore, and a medium unit at Rs 10 crore, with lower equipment thresholds for services. For the first time, a "medium" tier and a service-sector definition existed in law.

Second, and most consequentially for what follows, it built a delayed-payment machinery in Sections 15 to 24. Section 15 obliges a buyer to pay a micro or small supplier by the agreed date, and its proviso hard-caps any written agreement at forty-five days from acceptance, so no contract, however negotiated, can lawfully push payment beyond that line. Section 16 makes a defaulting buyer liable, notwithstanding any agreement to the contrary, to compound interest with monthly rests at three times the Reserve Bank of India's notified bank rate. Sections 18 to 19 route disputes to a state-level Micro and Small Enterprise Facilitation Council for conciliation and then arbitration, and require any buyer challenging an award to first deposit seventy-five percent of it in court. Sections 22 to 24 add disclosure and override teeth: audited buyers must report unpaid MSME dues in their accounts, that interest cannot be claimed as a tax deduction, and the chapter overrides conflicting law.

Third, it established a National Board for Micro, Small and Medium Enterprises to review policy, and required every state to constitute Facilitation Councils. On paper, this was a strong regime. Whether it worked is a separate question, and the answer is the reason the 2026 amendment exists.

Two decades of redefinition

The definition the Act hard-coded in 2006 proved the part that aged fastest, and successive governments spent fifteen years reworking it. In September 2015 the Ministry replaced paper filing with the Udyog Aadhaar Memorandum, a single self-declared online form. An attempt to raise the investment ceilings by amendment was withdrawn in 2018, and a more ambitious bill to switch to a turnover-only test lapsed with the dissolution of the sixteenth Lok Sabha in 2019.

The decisive change came in 2020, and it came in two stages that are worth separating because secondary accounts often blur them. On 13 May 2020, as part of the Atmanirbhar Bharat package, the Finance Minister announced a new composite definition using investment and turnover together, provisionally setting the medium-enterprise ceiling at Rs 20 crore of investment and Rs 100 crore of turnover. After representations that the turnover figure was too low, the Union Cabinet approved an upward revision on 1 June 2020, raising the medium ceiling to Rs 50 crore and Rs 250 crore. It was this second figure that was gazetted and took effect on 1 July 2020: micro up to Rs 1 crore and Rs 5 crore, small up to Rs 10 crore and Rs 50 crore, medium up to Rs 50 crore and Rs 250 crore, with the manufacturing and services distinction abolished and export turnover excluded.

The same day, the Udyam Registration portal went live, replacing Udyog Aadhaar with a free, paperless, self-declaration system. A Udyam Assist Platform followed in January 2023 to bring informal micro-enterprises without a GST registration into the formal fold for priority-sector lending. Registration grew steeply: from 0.79 crore enterprises at the end of 2021-22 to 7.83 crore by late February 2026, with trading the largest registered segment.

The most recent revision came in the 2025-26 Union Budget, which raised the classification limits to 2.5 times the investment and 2 times the turnover then in force, gazetted on 21 March 2025 and effective 1 April 2025. A medium enterprise can now hold up to Rs 125 crore of investment and Rs 500 crore of turnover. That the government could raise thresholds only by fresh notification each time, working around limits written into the Act's own text, is precisely the friction the 2026 amendment sets out to remove.

Did the payment law work? An independent reading

The clearest way to judge the 2006 delayed-payment regime is to ask whether MSMEs actually got paid on time under it. The measured record says the substantive rule was strong and the enforcement around it was not.

The scale of the problem was first quantified in 2022, when a report by the Global Alliance for Mass Entrepreneurship and Dun and Bradstreet India estimated that Rs 10.7 lakh crore was locked up annually as payments owed to MSME suppliers, roughly six percent of India's gross value added. A continued tracking series run with FISME and C2FO found the stock had fallen to Rs 7.34 lakh crore by March 2024, still above four and a half percent of gross value added. The government's own Economic Survey 2025-26 put the figure at about Rs 8.1 lakh crore. These are different methodologies and dates rather than a contradiction, but every reading agrees the sum is very large and unresolved. The 2025 tracking found that nearly forty percent of outstanding delayed payments were owed by government agencies and public-sector undertakings, the buyer category the state itself controls.

The formal remedy performed poorly against that backdrop. The MSME Samadhaan portal, opened in October 2017 to let suppliers file grievances, had received 2,56,892 applications worth Rs 55,244 crore, of which the Ministry's own June 2026 backgrounder recorded only 58,148 disposed by Facilitation Councils, a cumulative disposal rate near twenty-three percent. A portal snapshot at the end of 2025 showed 52,744 applications, worth Rs 8,397 crore, that had not even been examined. Reported analysis of the annual flow suggests the disposal rate fell from roughly a third of complaints in 2020-21 to about four percent in 2025-26, which, if borne out, means the live backlog was compounding rather than clearing.

Two structural facts explain the gap between a strong rule and a weak result. The first is enforcement design. Under the original Section 19, a buyer who lost at a Facilitation Council could stall by depositing seventy-five percent of the award in court and challenging it, with no statutory deadline forcing a decision, so winning at the Council did not translate into being paid. The second is that most suppliers never filed at all. The Economic Survey observed that MSMEs frequently avoid Facilitation Council cases for fear of straining the commercial relationship and losing future orders, and the burden fell hardest on the smallest firms, whose payment delays ran materially longer than those of larger enterprises within the category. The rule was real. The power to enforce it, for a small supplier who depended on the buyer, was not.

The 2024 tax lever, and what it revealed about enforcement

The most instructive recent precedent for whether a payment mandate can move buyer behavior is not in the MSMED Act at all. It is Section 43B(h) of the Income Tax Act, inserted by the Finance Act 2023 and effective from 1 April 2024. The provision converted the forty-five-day rule from a civil remedy a supplier had to pursue into an automatic tax cost for the buyer: a sum owed to a registered micro or small enterprise beyond the deadline could be claimed as a deductible expense only in the year it was actually paid.

The evidence on its effect is genuinely two-sided, and that is what makes it useful rather than convenient. On one side, the delayed-payment stock tracked by the industry series fell by roughly thirty percent between the 2022 baseline and March 2024, and the Chief Economic Advisor credited part of that decline to Section 43B(h) working alongside growth in invoice discounting. On the other, the same period produced documented evasion: a garment-industry body projected thousands of crores of quarter losses and reported that some retailers cancelled orders with registered suppliers, and trade press described buyers shifting purchases to unregistered vendors or pressuring small suppliers to surrender their Udyam registration so the rule would no longer bite. The government pushed back with its own data, stating that only about 0.6 percent of Udyam registrations had been cancelled, for reasons recorded as ownership changes and duplicates rather than flight from the rule.

The lesson the record actually supports is narrow. A well-designed mandate can shift aggregate behavior, and it can also invite avoidance at the margins, and both can be true at once. That tension is the right lens for reading what the 2026 amendment is likely to do.

What the 2026 amendment enacts

The Micro, Small and Medium Enterprises Development (Amendment) Bill, 2026, Bill Number LXXII of 2026, was introduced in the Rajya Sabha on 28 July 2026 by Jitan Ram Manjhi, the Union Minister for Micro, Small and Medium Enterprises. The Rajya Sabha passed it by voice vote on 3 August, and the Lok Sabha passed it on 7 August with little debate amid opposition protests over unrelated matters. No referral to a parliamentary committee has been reported, and the Bill moved from introduction to passage in eleven days. Its provisions fall into four groups.

What it could change, and what it probably will not

The following reads the likely effects. It is analysis, not measurement: the amendment's provisions have not yet operated, and every forward statement here should be read as a projection grounded in the prior record rather than an observed result.

The fifty-percent release rule is the provision most likely to bite, because it attacks the specific mechanism that failed. For twenty years the weakness was not the award but the stall after it, and a court obligation to hand a supplier half of a disputed sum after six months changes the arithmetic of stalling. Its reach is bounded by the same pre-deposit gate that precedes it and by court capacity, but the design is aimed squarely at the real defect.

The Central Public Sector Enterprise TReDS mandate is promising in principle and constrained in two ways the record makes clear. Adoption of discounting platforms has been thin, with roughly one lakh MSMEs onboarded against a receivables gap that independent analysis places in the tens of lakhs of crores, and an earlier mandate on large private buyers saw only about a third of eligible firms register years after it took effect. A mandate on central public-sector buyers, who the state directly controls, is more enforceable than one on private firms, which is likely why it was drawn narrowly. But that same narrowness is the provision's ceiling: the largest category of late payers in the grievance data is private and institutional buyers, and they are left outside it.

The ninety-day mediation cap and the faster arbitration timeline address a genuine problem, and their effect depends entirely on capacity that does not yet exist. Facilitation Councils were disposing of a falling share of cases even before the amendment added new deadlines, and a statutory clock without more Councils, staff and digital systems risks setting targets that cannot be met. The amendment's own instruction to states to run an adequate number of Councils is an acknowledgment of the gap; whether states resource it is the variable that will decide the outcome.

Independent commentary has landed in a consistent place. FISME's secretary general called the amendments a positive step but not a complete solution, addressing three limited issues while leaving broader structural challenges untouched, and identified execution, not legal text, as the recurring weakness. The Vidhi Centre for Legal Policy welcomed the new Development Commissioner interface and the TReDS provision while noting their value depends on staffing and empowerment at the state level. Two further limits are worth naming plainly: the forty-five-day protection applies only to micro and small enterprises, not medium ones, which can give an owner reason to cap growth rather than cross the threshold, cutting against the reform's stated aim; and buyer avoidance of the kind Section 43B(h) provoked remains a live possibility that stronger enforcement could invite as much as compliance.

The verdict

The 2026 amendment is best described precisely rather than generously. It is an enforcement-and-administration reform, not a change to the substantive rights the 2006 Act created. It leaves the forty-five-day rule and the interest penalty exactly where they were, and instead rebuilds the machinery that was supposed to make them real: a faster path through mediation and arbitration, a coercive route to recover an award, a rule that forces interim payment when a buyer stalls, a lighter and more digital registration and penalty regime, and a definition finally freed from the statute's own text.

Its most defensible feature is that it turns the state's discipline inward first, mandating discounting for the public-sector buyers who owe a large share of the arrears, rather than announcing a rule for private firms it cannot easily enforce. Its clearest limit is the mirror image of that choice: the private and institutional buyers who dominate the grievance data sit outside the central payments mandate, and the Facilitation Council capacity on which the new deadlines depend has been shrinking, not growing.

Whether the amendment moves the roughly Rs 8 lakh crore that the government's own survey says remains locked in delayed payments is therefore not answerable from the text. It will be decided by execution, by how many Councils states actually staff, by how far the discounting mandate is extended beyond central enterprises, and by whether stronger enforcement produces payment or avoidance. The law has been improved where it most visibly failed. The harder problems it was written to solve, the imbalance of power between a small supplier and a large buyer, and the distance between an award and a bank credit, are narrowed by this Bill, not closed.