How Indian Is India’s Export Boom?
India’s domestic value embodied in exports rose by $211.8 billion in current-dollar OECD estimates between 2015 and 2022, while its share of gross exports fell by 6.08 percentage points. The pairing shows why neither the export headline nor a local-content percentage can stand alone.
In 2020, India’s exports became more domestic by one familiar yardstick. The share of their value generated inside the country rose from 78.4% to 81.6%. Yet the amount generated in India fell by $4.5 billion. The more domestic export basket was also the smaller one.
The apparent contradiction in the OECD’s Trade in Value Added estimates is arithmetic, not error. A share can rise because its denominator contracts faster than its numerator. In this case, gross exports fell more sharply than the domestic value embodied in them. A higher domestic-content percentage therefore accompanied a decline in domestic value measured in current dollars. [1], [2]
The longer comparison runs in the opposite direction. From 2015 to 2022, India’s gross exports in the same dataset rose from $413.0 billion to $731.8 billion. Domestic value embodied in those exports rose from $332.1 billion to $543.8 billion: an increase of almost 64% in current dollars. But its share of the export total fell from 80.4% to 74.3%. [1]
A rising share did not prevent the 2020 fall in domestic value, and a falling share did not prevent the 2015–2022 increase. The useful question is how much additional value India generated through exports, and what changed the relationship between that value and the gross export bill.
The part of an export that was produced in India
A gross export records the value of the product or service crossing from a resident supplier to a non-resident customer. It can include value previously created abroad: a component, a chemical input or a business service used somewhere upstream. Domestic value added traces the part generated within India, including contributions from domestic suppliers to the exporting industry. The OECD’s input-output accounts connect these stages rather than treating the exporter’s final invoice as entirely new Indian production. [2]
“Domestic” describes the location of value generation. It does not mean Indian ownership. A foreign-owned establishment operating in India can generate domestic value; an Indian-owned establishment producing abroad does not become domestic to India merely because of its owner. Nor does the measure tell us how the value is divided between wages, profits and other components, or how much profit is retained in the country. [2]
Consider a deliberately simple illustration. An export worth 100 contains 80 of domestic value. If a different production arrangement raises the export to 200 and domestic value to 120, the domestic share falls from 80% to 60% while the domestic amount rises by half. The illustration identifies an arithmetic possibility. Whether access to imported inputs actually causes such an expansion requires separate evidence.
Of the $318.7 billion increase in gross exports between the two endpoints, $211.8 billion—66.4%—is accounted for arithmetically by higher domestic value added. Foreign value added accounts for about $106.9 billion. Tiny discrepancies reflect the precision of the downloaded components. These are parts of an accounting change, not estimates of how much growth imports caused. [1]
The foreign component grew faster, by about 132%, than the domestic component. That is why its share increased. It does not follow that India lost $106.9 billion that it could otherwise have produced. Producing the same export basket without those foreign contributions might involve different costs, quantities and products. The observed accounts do not provide that counterfactual.
The changing basket is not the whole explanation
A falling aggregate domestic share could arise even if every industry’s own ratio stayed unchanged. Suppose export growth shifted towards industries that had always used more foreign inputs. Their growing weight would pull down the national ratio. That explanation matters because it describes a changing export mix, rather than a change inside any one broad industry.
To distinguish the two, we divide exports into seven non-overlapping groups: agriculture; mining; manufacturing; electricity and related supply; water and waste services; construction; and the remaining services. We then separate the 6.08 percentage-point fall into changes in the groups’ own domestic-value ratios and changes in their export weights. The calculation gives each component the average of its beginning- and end-period weight or ratio, so the two terms add back to the observed change. [1], [2]
Changes in the ratios within these broad groups account for 5.72 percentage points of the fall. Changes in export composition account for 0.36 points. In this partition, roughly 94% of the decline lies within the seven groups. A still coarser split between goods, utilities and construction on one side and services on the other produces a similar result: 5.65 points within the groups and 0.42 points from composition. [1]
This rules out an explanation based entirely on shifts between these broad categories. It does not establish that individual factories substituted imported inputs for Indian ones. Each category contains many products and firms; its internal mix can change. Relative prices can also change the ratio. The decomposition locates the accounting movement at a chosen level of aggregation. It does not identify its cause.
Manufacturing supplies the largest within-group contribution, about 4.49 percentage points of the national fall. Its own domestic share declined from 71.0% to 62.8%. Services supply another 1.22 points of the within-group component, although their domestic share remained much higher, falling from 91.6% to 88.7%. [1]
An obvious concern is petroleum. Changes in the value of imported energy can move nominal input-output ratios without an equivalent change in production technology. Excluding coke and refined petroleum from manufacturing leaves a 6.19-point fall in the remaining manufacturing ratio. Excluding that industry from the entire export basket leaves a 4.42-point decline. The direction therefore survives that exclusion. It does not remove energy inputs used elsewhere or turn current-dollar observations into volume measures. [1], [2]
The computers, electronic and optical products category makes the levels-versus-shares distinction especially clear. Its domestic value added rose from $5.8 billion to $16.7 billion, while its domestic share fell from 71.9% to 58.2%. This broad industry category does not isolate phones, a named company or a subsidy programme. Its result cannot settle their performance. It does show why a lower share alone cannot establish that domestic production failed to expand. [1]
The gains look different from the gross headline
The gross export increase was larger in manufacturing than in services. Manufacturing exports rose by $184.4 billion, compared with $124.6 billion for services. Looking only at those increases would place manufacturing at the centre of the story. [1]
The domestic-value comparison changes their relative importance. Domestic value embodied in manufacturing exports rose by $97.5 billion. Domestic value embodied in services exports rose by $105.3 billion. Services therefore account for about 49.7% of the total domestic-value increase, and manufacturing for 46.1%; the other broad sectors supply the remainder. [1]
This classifies value by the industry exporting it, not necessarily by the industry that originally generated every part of it. Domestic services supplied to a manufacturer can be embodied in manufacturing exports. The two figures cannot be relabelled as the manufacturing and services sectors’ contributions to GDP growth. [2]
They nevertheless change the comparison that an export headline invites. Manufacturing produced the larger rise in gross exports, while services exports embodied the slightly larger rise in domestic value. The distinction matters when deciding which activities deserve further investigation: the most visible shipments need not tell the whole story of value generated at home.
Imported inputs can be part of the capability
The strongest rival to a pessimistic reading of the declining share is that imported inputs can make domestic production possible or more competitive. Access to a component, material or specialised service can expand what a firm is able to sell. More imported content and more domestic activity can then occur together.
There is historical Indian evidence for this mechanism. Goldberg, Khandelwal, Pavcnik and Topalova studied relatively large manufacturing firms around India’s earlier trade liberalisation. In the NBER working-paper version examined here, they connect lower input tariffs with an expansion in firms’ product range, including access to previously unavailable intermediate inputs. Their sample covers 1989–1997, and results vary across specifications; the long-difference estimate is not conventionally statistically significant. This is evidence about a mechanism in a particular reform episode, not proof of what caused the 2015–2022 aggregate change. [3]
Veeramani and Dhir’s later study also directly addresses the distinction between domestic shares and domestic levels in Indian production networks. The accessible publisher abstract reports gains in export, domestic-value and employment levels with greater participation in global value chains. Its older national input-output sample and identification assumptions differ from the OECD observations used here; the full published empirical body was not accessible for this investigation. It supplies a relevant competing interpretation, not a coefficient that can be imported into our calculation. [4]
For any input, the policy question is concrete: can a locally supplied alternative match or improve the relevant combination of cost, quality, reliability and capability? If it can, deeper domestic sourcing may raise the amount generated at home without sacrificing the export opportunity. If it cannot, demanding a higher local share may shrink the activity to which the percentage applies. These accounts distinguish the quantities a policy should examine; they do not select the answer for a particular industry.
Where the reconstruction stops
The observations are annual, current-dollar OECD estimates for 2015–2022, frozen from the download of 5 September 2026. They are not a new survey of Indian firms or inflation-adjusted measures of productive capability. OECD’s input-output construction reconciles several statistical sources and can extrapolate structures where newer detailed tables are unavailable. Reproducing the calculations verifies their arithmetic, not the truth of every underlying estimate. [1], [2]
Newer OECD chain-linked indicators exist, including projected 2023–2024 observations. Their supporting access was inadequate for this analysis, so those years have not been appended to this series. A general consumer-price deflator would not resolve industry-specific relative-price changes or make the measures interchangeable. Claims about real upgrading, employment, welfare or the returns to an incentive programme remain outside this reconstruction. [5]
Within those limits, four results remain. Domestic value grew substantially in current dollars, while its share fell. Most of that share decline sits within broad industry groups rather than between them. Services and manufacturing contribute differently depending on whether the comparison is gross exports or domestic value. And 2020 demonstrates, within the same series, that a more domestic percentage can accompany less domestic value.
The next decisive evidence would separate price changes, finer product composition and production relationships inside those broad groups. Until then, the object to explain is the combination: why the domestic amount increased while the domestic share fell—and under what conditions India could raise the former further. Raising the percentage alone is not an adequate definition of success.
Sources and calculation notes
[1] OECD, Trade in Value Added, flow OECD.STI.PIE:DSD_TIVA_MAINLV@DF_MAINLV(1.1), India to World, annual 2015–2022, EXGR, EXGR_DVA and EXGR_FVA, current USD millions. Frozen retrieval 5 September 2026, 15:13:42 UTC. Calculations use the saved raw response, not an assumption that a flow version is immutable. Original data request.
[2] OECD, Guide to Trade in Value Added Indicators, 2025 edition: units p.5; exports p.16; domestic value and ratios p.19; foreign value p.23; estimation p.43; classification pp.46–47. Original guide. The raw 79 activity codes contain overlapping aggregates and are never summed together. The seven-group partition uses A/B/C/D/E/F/GTT. The midpoint decomposition is ΔS = Σ(mean weight × change in ratio) + Σ(mean ratio × change in weight), with share changes expressed in percentage points.
[3] Goldberg, P., Khandelwal, A., Pavcnik, N. and Topalova, P., Imported Intermediate Inputs and Domestic Product Growth: Evidence from India. NBER Working Paper 14416 (October 2008); the exact downloaded PDF says “revised September 2009” internally, while the NBER landing page lists a December 2011 revision. Tables 4a/4b and discussion were inspected; this is not an assertion that the final published tables were inspected. NBER record.
[4] Veeramani, C. and Dhir, G. (2022), “Do developing countries gain by participating in global value chains? Evidence from India,” Review of World Economics 158, 1011–1042, DOI 10.1007/s10290-021-00452-z. Publisher abstract, notes and instrument appendix accessible; the full published empirical body was not inspected. Publisher.
[5] OECD, “Global value chain data across trade, multinationals and services.” The original description identifies newer chain-linked measures and projected years. No newer observations are incorporated here. Dataset description.