Make in India needs faster signals: What electricity reveals about manufacturing
More than a decade after the launch of Make in India, manufacturing remains central to India’s growth strategy. The ambition-to raise manufacturing’s share of GDP from about 15-16% to 25%-has been reiterated across policy documents, industrial corridors, and Production Linked Incentive (PLI) schemes spanning fourteen sectors. Billions in fiscal support and regulatory effort are tied to this goal. Yet a striking irony remains: India continues to measure manufacturing output far too slowly.
The most comprehensive manufacturing dataset, the Annual Survey of Industries (ASI), is released with a lag of 18-24 months. Manufacturing Gross Value Added (GVA) in the National Accounts Statistics (NAS) becomes available about a year after the reference period and undergoes multiple revisions. State-level manufacturing GVA appears with even longer delays. The Index of Industrial Production (IIP), while monthly, is an index with limited coverage, high volatility, and well-known measurement issues. For an economy aspiring to global manufacturing competitiveness, this constitutes a serious information gap.
At the same time, India already tracks a variable at high frequency, across states and sectors, with minimal reporting error: electricity consumption. The question is no longer whether electricity can proxy manufacturing activity, but whether India is willing to use it systematically.
Why electricity is a meaningful signal
Manufacturing is energy-intensive. Electricity powers machines and assembly lines. Unlike labour or capacity utilisation (difficult to observe directly), electricity use is metered, continuous, and hard to misreport, and is available monthly (often daily) and at fine geographic and sectoral granularity.
This intuition is supported by international evidence. During the Covid-19 pandemic, European studies using high-frequency electricity market data showed that power demand closely tracked economic contraction and recovery when official output data were delayed.
In China, sector-level electricity consumption mirrored factory shutdowns during lockdowns and rebounded early during reopening, signalling manufacturing recovery ahead of official statistics. Evidence from developing economies suggests a strong long-run relationship: panel studies for West Africa, including Nigeria, find a statistically significant association between electricity consumption and factory-based industrial output.
What India’s own data show
Indian data strongly echo these findings. Over the past fifteen years, all-India manufacturing GVA and electricity consumption move almost in lockstep (correlation = 0.99). This relationship holds across policy regimes, business cycles, and even the pandemic disruption.
Moving from correlation to real-time estimation of manufacturing activity would require an explicit econometric framework. Electricity consumption from industrial feeders and high-tension manufacturing connections, already recorded by state electricity distribution companies (DISCOMs), can be mapped to National Industrial Classification (NIC) industries and aggregated weekly or monthly at the state and sector level.
With appropriate controls for scale, industrial structure, and long-run trends, an econometric model can be developed that can generate early signals of manufacturing activity well before official GVA estimates are released.
Where structure matters
The relationship also holds below the national level. For instance, Gujarat, a high manufacturing-led economy with relatively reliable power supply and dense industrial clusters, represents a textbook case for electricity-based monitoring, with correlations close to 0.96.
The relationship is particularly strong (around 0.9) in other manufacturing-led economies such as Maharashtra, Karnataka, and Uttar Pradesh.
Where correlations weaken, the reasons are structural rather than statistical. Delhi’s economy is services-dominated; smaller UTs and tourism-dependent regions have limited industrial bases.
This heterogeneity is informative: electricity works best as a monitoring tool where manufacturing is economically significant.
Sectoral evidence: where electricity works best
Sector-level analysis provides the strongest validation. Food processing is employment-intensive, closely linked to agriculture, and central to inflation management and exports. In this sector, electricity consumption tracks manufacturing output almost perfectly over time, including during the post-pandemic recovery. Under Make in India, the PLI scheme allocates ?10,900 crore to food processing to modernise capacity and boost output in ready-to-eat foods, processed fruits and vegetables, marine products, and more-precisely the segments where electricity data can provide nimble monitoring. For an agenda that prioritises jobs and value addition, this is a powerful insight.
By contrast, electricity performs poorly in activities such as repair and installation services or certain petroleum-related industries, where output is price-driven or service-like.
What should change
Using electricity data as a real-time proxy for manufacturing activity will require a coordinated institutional effort. MoSPI should anchor this process by constiting an expert group to design the measurement framework, including data architecture and indicator protocols. This effort must involve the Central Electricity Authority (CEA) and state DISCOMs, which already collect detailed consumption data but maintain databases not designed for statistical integration. Standardised instructions from the CEA to DISCOMs would enable systematic mapping of granular power data to industrial activity (by NIC code) and comparability across states. Additional support is needed for states to link electricity connection data with administrative factory records maintained by the Chief Inspector of Factories. Over time, these linkages could strengthen the ASI sampling frame, improving coverage of smaller and underrepresented units, especially MSMEs.
The future
The transition to cleaner energy and rising efficiency will alter electricity intensity over time. Yet, even with efficiency gains, electricity demand continues to track industrial cycles. Direction and timing often matter more than levels. Make in India is about factories, jobs, and competitiveness. Measuring manufacturing output two years after the fact is a luxury India can no longer afford.
Ashish Kumar is President of the Center of Data for Economic Decision-making and Chief Statistician at Pahle India Foundation; Former Director General, MOSPI. Payal Seth is Head of CoDED and Lead Economist at Pahle India Foundation ; views are personal
