Is it in the interest of a country to promote strong local linkages for its domestic industries or to participate in global production sharing activities wherein linkages are globally dispersed? This article informs this debate by empirically analysing which one of these strategies would result in higher levels of domestic value added and employment in a developing country like India.
Does export of goods offer a viable path to job creation? While this is an important question for policymakers, the answer is not as straightforward as it may sound. The relationship between exports and employment becomes complex in the context of global production sharing (GPS) activities that are entrenched and ubiquitous in today’s world. The phenomenon of globally fragmented production processes, often described by the term GPS, implies that intermediate inputs cross borders several times during the manufacturing process. This leads to double (or multiple) counting as trade data are usually collected and reported as gross flows at each border crossing rather than net value added between border crossings. In other words, official trade data does not capture the domestic value added (DVA) content of exports. Yet, DVA is what really matters for job creation within a country. Thus, a basic requirement for answering the above question is the availability of data on the DVA content of exports.
Another issue is conceptual: the mechanisms underlying the relationship between exports and jobs are no longer straightforward when a country participates in GPS. A practical question for policy is whether it is desirable to promote strong local linkages for domestic industries by sourcing intermediate inputs domestically or to participate in GPS wherein linkages are globally dispersed. Our recent research (Veeramani and Dhir 2019) informs this debate by empirically analysing which one of these strategies would result in higher levels of DVA and employment in India.
Indeed, participation in GPS, as opposed to local sourcing of intermediate inputs, implies that the ratio of DVA to gross exports (henceforth DVAX ratio) is much smaller. However, driven by the scale and productivity effects of producing for the world market, the absolute levels of DVA and job creation could be much higher. To give an example, Dedrick et al. (2010) estimated that the factory-gate price of an assembled iPod from a Chinese factory was US$144, but only about US$4 of this constituted Chinese value added with much of the rest being captured by the US, Japan, and Korea. However, despite the low DVAX ratio, the absolute value of Chinese DVA from iPod assembly is very high due to the scale effect1.
Using Input-Output (IO) analysis2, in Veeramani and Dhir (2019), we build a consistent panel dataset on DVA and jobs tied to Indian exports from 112 sectors, covering the whole economy, for the period from 1999-2000 to 2012-2013. The major advantage of the IO framework is that it enables us to disentangle the direct and indirect effects (backward and forward linkages) of exports from any given sector. For example, exports of ‘automobiles’ generates direct DVA within the automobile sector as well as in other upstream sectors, such as ‘iron and steel’, whose outputs are used as inputs by the automobile sector. The DVAX ratio can be used as a proxy to quantify the extent of a country’s participation in GPS (Johnson and Noguera 2012), the rationale being that this ratio measures how much domestic, as opposed to foreign, value-added is generated throughout the economy for a given unit of exports. In general, sectors with greater participation in GPS tend to record lower DVAX ratio and vice versa. We carry out a regression analysis to test the hypothesis that greater participation in GPS leads to higher absolute levels of gross exports, DVA, and employment.
Findings and discussion
The DVA content of India’s aggregate exports (merchandise plus services) increased from US$46 billion in 1999-00 to US$295 billion in 2012-13, with a growth rate of 17.7% per annum. During this period, the total number of direct plus indirect jobs supported by aggregate Indian exports increased from about 34 million to 62.6 million, with a growth rate of 3.4% per annum. Throughout the period, export related jobs grew faster than that of total employment – the share of export-supported jobs in total employment increased from little over 9% in 1999-00 to 14.5% in 2012-13. Export supported jobs accounted for 39.5% of total employment in the manufacturing sector in 2012-13, up from 19.6% in 1999-00. Backward linkages, particularly from manufacturing to agriculture and services, have become an important source of export related DVA and job creation in the country.
The DVAX ratio steadily declined from 0.86 in 1999-00 to 0.65 in 2012-13, with the decline being particularly sharp for the manufacturing sector during the second half of the 2000s. These trends suggest that Indian industries have increased their participation in GPS over the years. The econometric analysis confirms that a greater participation in GPS leads to higher absolute levels of gross exports, DVA, and employment (see Veeramani and Dhir 2019 for details). This result is robust to a variety of specifications and implies that the positive scale and productivity effects of participation in GPS outweigh any possible negative impact of the same.
A pertinent question, in the light of the above findings is: why has the manufacturing sector, despite its increased participation in GPS, not yet become the engine of India’s growth unlike for China and other major East and South-East Asian countries? While this question begs a detailed analysis, an important point to note at this juncture is that while India’s involvement in GPS has indeed increased over the years, the level of its participation remains significantly less than that of other countries in East Asia (Veeramani and Dhir 2017). The Trade in Value Added (TiVA) database, maintained by the Organisation for Economic Co-operation and Development (OECD), shows that, for the year 2011, the DVAX ratio for India’s manufacturing sector was 0.64 as compared to 0.48 for Malaysia, 0.51 for Singapore and Vietnam, 0.52 for Thailand, 0.53 for Korea, and 0.60 for China. The difference between India and other countries is starker for sectors, such as electronics and electrical machinery, where GPS activities are generally more prevalent3.
India can reap rich dividends by adopting policies aimed at strengthening its involvement in GPS. There are unexploited potential in India’s traditional labour-intensive industries such as textiles, clothing, and footwear. Further, based on imported parts and components, India has a potential to emerge as a major hub for final assembly in several industries, particularly in electronics and electrical machinery. While global value chains in several industries are primarily controlled by big multinational enterprises (MNEs), local firms play a role as subcontractors and suppliers of intermediate inputs to MNEs. For example, firms in auto ancillary industries supply components to larger manufacturers.
For a country to become an attractive location for assembly activities, it is important that import tariff rates for intermediate inputs are zero or negligible. The move by Indian government, since late 2017, to increase import duties for some of the intermediate inputs, partly in retaliation to the recent US tariff hikes and partly to boost the ‘Make in India’ initiative is not a move in the right direction. The import-weighted average tariff rates for intermediate inputs, for the year 2017, already stood high at 9.23% for India compared to 4.41% for China.
It is important to create an ecosystem which will result in a realignment of India’s specialisation patterns towards labour-intensive processes and product lines. A number of studies have noted an idiosyncratic nature of India’s specialisation patterns in that, despite being a labour-abundant country, the fast growing exports are either capital-intensive or skill-intensive (Kochhar et al. 2006, Panagariya 2008, Veeramani et al. 2017). Some authors have argued that India’s labour laws create severe exit barriers and hence discourage large manufacturing from choosing labour-intensive activities and technologies (Panagariya 2008).
Studies show that a low level of service link costs – costs related to transportation, communication, and other related tasks involved in coordinating the activity in a given country with what is done in other countries within the production network – is critical for countries to involve in GPS. Supply disruption in a given location due to shipping delays, power failure, political disturbances, labour disputes, etc., could disrupt the entire production chain. Finally, participation in GVCs requires not only trainable low-cost unskilled labour but also a lot of middle-level supervisory manpower. Concerted efforts are needed to improve the quality of education at all levels, to expand vocational training, and to facilitate large scale apprenticeship programmes.
- Consider this simple back-of-the-envelope calculation. In 2008 (close to the years for which Dedrick et al. provided the estimates) Apple sold 54.83 million units of iPods. Assuming that the whole assembly was done in China, the aggregate Chinese DVA from the assembly of this single product was 219 million dollars ($4 × 54.83 million units), not a trivial amount.
- An IO table is a matrix whose entries represent the transactions occurring between sectors (the output of a sector is the input of another). Each row in the table shows how a given sector’s output is used as input by various sectors (forward linkage) while each column records a given sector’s purchase of inputs from other sectors (backward linkage).
- It may be noted that the DVAX ratio reported in TiVA is an overestimation for countries heavily involved in processing trade (such as China and Mexico) because of the underlying assumption that production techniques and input requirements are identical for exports and domestically absorbed final goods (Koopman et al. 2010, Johnson and Noguera 2012). Processing exports account for about half of overall exports in China. Johnson and Noguera (2012) showed that, once processing exports are separately taken into account, the DVAX ratio for the year 2004 fell from 0.70 to 0.59 for China, and from 0.67 to 0.52 for Mexico.
- Dedrick, Jason, Kenneth L Kraemer and Greg Linden (2010), “Who profits from innovation in global value chains? A study of the iPod and notebook PCs”, Industrial and Corporate Change, 19(1):81-116.
- Johnson, Robert C and Guillermo Noguera (2012), “Accounting for intermediates: production sharing and trade in value added”, Journal of International Economics, 86(2):224-236.
- Kochhar, Kalpana, Utsav Kumar, Raghuram Rajan, Arvind Subramanian and Ioannis Tokatlidis (2006), “India’s pattern of development: what happened, what follows?”, Journal of Monetary Economics, 53(5):981-1019.
- Koopman, R, W Powers, Z Wang and S Wei (2010), ‘Give credit where credit is due: tracing value added in global production chains’, NBER Working Paper Series, Working Paper No. 16426.
- Panagariya, A (2008), India: The Emerging Giant, Oxford University Press, New York.
- Veeramani, C and G Dhir (2017), ‘Make what in India?’, in S. Mahendra Dev (ed.), India Development Report 2017, Oxford University Press, New Delhi.
- Veeramani, C, Aerath Lakshmi and Prachi Gupta (2018), “Intensive and extensive margins of exports: what can India learn from China?”, The World Economy, 41(5):1196-1222.
- Veeramani, C and G Dhir (2019), ‘Reaping Gains from Global Production Sharing: Domestic Value Addition and Job Creation by Indian Exports’, IGIDR Working Paper WP-2019-024, Indira Gandhi Institute of Development Research, Mumbai.