Livelihoods or Enterprises? What Budget 2026–27 Reveals About India’s Enterprise Reality

What Budget 2026-27 gets right about India’s enterprise landscape — and what it still can’t see

Every year, we have the ritual of the budget and economic survey. Rituals are important because they allow time to introspect and plan ahead. The story that emerges from this year’s economic survey and budget is of a vision of India, in the near future, a few decades ahead, as an industry-led developed country. It is not a full-fledged dream of abundance, but of selective, cautious ambition. The economic survey and the budget have identified, largely in an implicit way, the more organised, larger enterprises as the engine that should fire this dream. And that’s not wrong.

Budget 2026-27’s flagship enterprise provisions tell this story clearly. The ₹10,000 crore SME Growth Fund targets “Future Champions” based on export readiness, technical capability, and scaling potential. The TReDS architecture, now mandated for all central PSU purchases from MSMEs, presupposes enterprises already embedded in government procurement. The Self-Reliant India Fund’s ₹2,000 crore top-up extends equity to MSMEs “with potential and viability to grow into large units.” These are coherent interventions for businesses that can absorb equity, integrate into formal supply chains, and achieve scale. At roughly 15% of GDP, government expenditure cannot efficiently support all economic activity — selectivity is rational.

It is also rational because many at the lower end are labelled ‘enterprises’ by definitional fiat rather than self-identification. But the budget’s near-silence on the lower end raises a question that the Economic Survey’s treatment of MUDRA and SVANidhi makes vivid: are we supporting livelihoods or enterprises? The Survey documents 20% business income growth among SVANidhi-supported street vendors and interprets this through financial inclusion. Yet measurable enterprise performance gains are evident in the same data. This ambiguity persists because we lack the vocabulary — and the numbers — to distinguish between the two.

The riddle of enterprise numbers: ASUSE 2023-24 estimated 7.34 crore unincorporated enterprises — proprietary, partnerships, and collective — excluding construction. In September 2024, there were about 5.25 crore Udyam and Udyam Assist Platform-registered enterprises. As argued here, these two universes have little in common. ASUSE reports only 0.5% of its enterprises with Udyam registration; 63% report not being registered under any act or authority. If we assume these are two separate sets, total enterprise employment reaches 35 crore — a number that does not sit well with the accepted understanding of India’s employment structure.

Table 1: Udyam-Registered vs Non-Registered Enterprises (ASUSE 2023-24)

Source: ASUSE 2023-24 microdata analysis; weighted estimates.

But the riddle goes deeper than the ASUSE-Udyam mismatch. Even within the ASUSE universe, the 7.34 crore number bundles together activities of fundamentally different character. Using ASUSE 2023-24 microdata, we can decompose this universe.

At one end are activities that are clearly not enterprises in any conventional sense: 63 lakh households earning rental income, 30 lakh financial intermediaries (predominantly chit fund operators with a median annual surplus of ₹5,600), and 13 lakh domestic workers serving multiple households. At the other end are 1.04 crore home-based manufacturing workers — overwhelmingly women (90%), operating from household premises, with a median annual turnover of ₹40,680. These are piece-rate workers in the putting-out system: the survey counts them because they are technically own-account, but they have no customers, no market presence, no autonomy over production decisions.

Between these extremes sit 26 lakh self-employed professionals — tutors, doctors, lawyers, consultants — and 66 lakh transport owner-operators. Whether these constitute ‘enterprises’ is a question the reader can assess; the auto-rickshaw driver owns a productive asset and faces market risk, but the solo tutor giving home lessons is essentially selling labour with minimal capital.

Table 2: The Enterprise Universe Decomposed (ASUSE 2023-24)

Source: ASUSE 2023-24 microdata analysis; weighted estimates.

Core enterprises = all activities not classified in other categories. HWE = Hired Worker Enterprise; OAE = Own Account Enterprise. Dependent home manufacturing = manufacturing OAE, operating from household premises, annual turnover below ₹1 lakh.

Strip away the non-enterprise activities and the ‘enterprise’ population shrinks from 7.34 crore to about 4.8 crore — and even this core is dominated by 3.8 crore own-account operators with a median turnover of ₹4 lakh and no hired workers. The enterprises that resemble what enterprise policy typically imagines — units that hire workers, operate from fixed premises, generate substantial turnover — number roughly 1 crore. This is still a large and important population, but it is a seventh of the official count, not the count itself.

The mirage of large numbers has real consequences. Scarce resources of the government and society — CSOs, philanthropy, development finance — are spread thin across livelihood actors misidentified as enterprises. The growing recognition, in social actors as well as in government, to focus on a subset of these official enterprises is welcome. Livelihood enterprises do not graduate into hiring workers, larger enterprises — a fact proven by data as much as anecdote. To make the most of value generation opportunities, enterprises need to begin at scale.

The decomposition table offers a clue to resolving the riddle: the single variable that most cleanly separates enterprise-like activity from other economic activity is whether the unit hires workers. Among core enterprises, those with at least one hired worker — roughly 1 crore units — have a median turnover 3.6 times higher, fixed assets 3 times larger, and are far more likely to access formal credit. They are the population that enterprise policy addresses. The remaining 3.8 crore core OAEs are self-employed operators: enterprising, often, but structurally closer to the livelihood activities that the budget — correctly — addresses through different instruments.

This is not a uniquely Indian insight. The global practice of defining ‘enterprise’ is typically through employment generation, separating self-employment — which can be enterprising but is often restricted to imitative service provision or retailing in a competitive market — from entrepreneurial activity involving substantial value creation. India has not adopted this lens, for a practical reason: employee numbers are likely to be underreported. The labour regulation regime, whatever its merits, creates incentives for employers to keep hiring relationships informal and invisible. Resolving this — through simple, non-threatening mechanisms for registering every hiring relationship, leveraging the digital public infrastructure we have built — would let us count enterprises the way the rest of the world does.

The economic survey and the budget have done the right thing, without explicitly stating so, by separating livelihood strategy for self-employment-driven activities from industrial policy for growth-oriented, functionally sophisticated enterprises. Making this separation explicit — in how we define, count, and target enterprises — would let both strategies work better.