Intermediate inputs – the parts and materials imported to make products for consumption domestically and abroad – are a growing force in world trade. Catalysed by the globalisation of production, the large and rapidly increasing use of imported inputs for exports has important policy implications. Bilateral trade balances are not appropriately measured, the costs of protection are higher than often understood, trade is more volatile, and the importance of exports as drivers of short-term demand is overestimated.
The rise of trade in intermediates
Lower trade barriers, organisational innovations, and progress in information and communication technologies have made slicing up the production process cheaper and easier. Coordination costs have fallen, and different stages of production are now more frequently located in different countries. High labour costs and heavy regulations in rich countries have also helped to accelerate the shift through a wave of outsourcing and offshoring to developing countries (Cappariello and Zizza 2007).
As a result, intermediate inputs have become a salient part of world trade, particularly as imports of these goods have increased sharply relative to their total use. Intermediate inputs now represent more than half of the goods imported by OECD economies and close to three-quarters of the imports of large developing economies, such as China and Brazil.
Most importantly, they account for a significant chunk of exports, with large differences across countries. According to OECD estimates1, imported content accounts for about a quarter of OECD economies’ exports, and the European Central Bank estimates that import content accounted for about 44% (or 20% for extra-EU imports) of EU exports in 2000, ranging from about 35% in Italy to about 59% in the Netherlands2. In the US, the import content of exports was about 10% in 2005. Among emerging economies, imported content’s share in exports is particularly high in China – about 30%, or twice that for India and Brazil.
Figure 1. Imported content of exports
Source: OECD IO Database
With globalisation, the use of imported intermediates for exports has been growing. According to the OECD, all but one of its member countries increased the import content of its exports over 1995-2005. The increase was particularly marked in small countries like Luxemburg and Israel, which saw increases of about 20 percentage points, compared to 3-8 percentage points in the large countries, such as the US, Japan, and Germany. This is in keeping with the general trend of import content accounting for a larger share of exports in smaller economies.
The growing role of imported intermediate inputs has several implications for economic study and therefore policy.
- The importance of bilateral trade balances is exaggerated
Economists have long argued that only overall – and not bilateral – trade balances matter. Acting on bilateral imbalances without addressing the underlying causes of the aggregate imbalance simply redistributes that imbalance across trading partners.
Now, with the role of trade in intermediates rising, bilateral trade balances are even less meaningful – they do not reflect value added. As WTO Director General Pascal Lamy argued recently, many countries’ exports, including those of China, are economically less significant than they look because so much consists of re-exports and the modest reprocessing of intermediates. Though each iPod touch contributes $150 to the US-China bilateral deficit, for instance, China adds only $4 to the value of each unit.
Various studies find that China’s surplus with the US, for example, is 20%-40% lower when estimated in value-added terms – reflecting the fact that its exports contain only 20%-35% of domestic value-added (see Maurer and Degain 2010). Japan’s and South Korea’s balances with the US, on the other hand, may be understated, since China relies on content imported from them to produce its exports. As they have exported more parts to China, Japan’s and South Korea’s share of US imports has declined.
Figure 2. US bilateral imports
Source: IMF direction of trade statistics
- The importance of exports as a driver of demand is overestimated, while the importance of trade as a source of efficiency is underestimated
Over the last several decades, world exports have grown at about twice the rate of world GDP on average. The increased trade in intermediate goods – commonly exported several times before becoming embodied in a final product – helps account for this, as shown by the fact that the sectors which have registered large export growth, such as machinery, are also the sectors where the most vertical specialisation growth has occurred (Nordås 2003).
The growth of trade in intermediate goods also helps explain why exports account for an enormous share of GDP in a few mega-traders, such as Singapore and Hong Kong, sometimes called entrepôt (or re-export) economies (O’Rourke 2009). Because policymakers fail to recognise that imported inputs feed into exports, they often overestimate the importance of exports as a driver of short-term demand but underestimate the importance of trade and specialisation as sources of increased efficiency in the longer term.
- Trade has become more volatile and a larger source of shocks
Generally, intermediate imports appear to be more important for exports of manufactures than those of services, particularly in industries such as electronic and communications equipment, and electrical machinery and instruments. In the US and Japan, the import content of manufactures exports – nearly 20% – is four times that of services exports; in China, it is twice that of services exports.
Figure 3. Import content of exports by industry
Source: OECD IO Database
At the same time, manufactures, especially durable goods, play a larger role in trade than in GDP – in the US, durables accounted for more than 60% of trade in goods in 2008, compared to 24% of GDP – but the demand for durable goods tends to fluctuate more than that for services. As a result, trade is more volatile than GDP, and the effect is compounded by the fact that durable goods account for a high share of trade in components.
The Great Recession provided a dramatic illustration of this. Global exports declined by 14% in volume terms between the third quarter of 2008 and first quarter of 2009, while world GDP declined by about 3% over the same period (see also Baldwin 2009). Not surprisingly, trade in capital and durable goods was hit particularly hard; according to an IMF study, during the worst of the crisis, it fell about 10 times faster than trade in consumer non-durables, as amid a global credit crunch and loss of confidence consumers postponed any purchases that could be delayed. In addition, due to countries’ specialisation in different stages of production, shocks in one country could forcefully translate to shocks to stages undertaken in another, magnifying the disruption.
Figure 4. The composition of trade volumes
Though such trade volatility does not necessarily translate into equivalent changes in domestic value-added, it is nonetheless highly disruptive. With trade in intermediates growing, economies are becoming more intertwined, implying greater vulnerability to shocks emanating from abroad. At the same time, increased reliance on foreign demand and supply is making economies less vulnerable to domestic shocks.
- The cost of protection is higher
Trade in intermediates means that the cost of protection is higher than is generally understood, and rising. As economists have long known, the effective rate of protection – the tariff as a share of domestic value-added – is higher than the nominal tariff. Consider, for example, a t-shirt produced in the US. Assume it trades at $10 and uses $5-worth of imported fabric. The domestic value-added is therefore $5. Now, if the US imposes a tariff of 50% on t-shirts, the price of an imported t-shirt will rise to $15, giving domestic industry a 100% price advantage3.
By the same token, levying a 50% tariff on the fabric imports would increase the costs for t-shirt exporters by 50% of their value-added – effectively creating an export tax. Because imports increasingly feed into exports, an import tariff on parts and raw materials has a big impact on exports. Tariffs on intermediates may also discourage inward bound foreign direct investment and encourage outward bound instead4.
The danger of higher protection is particularly pronounced for smaller economies where the share of intermediate imports in a country’s overall exports is large.
Figure 5. Import content of exports with partner countries
Source: OECD IO Database
In addition, higher trade barriers may be particularly disruptive to intra-regional trade, as countries tend to import intermediate inputs from other countries in their region, partly reflecting production networks’ high sensitivity to time constraints, trade, and transportation costs (Yi 2009). EU countries tend to import intermediates from other EU members, NAFTA countries from other NAFTA countries, and Japan, China, Korea, and Indonesia from other countries in Asia (see chart above).
It is important to develop better measures of trade flows net of intermediate imports. A failure to do so can lead to the wrong policy conclusions about the importance of bilateral trade imbalances, and can lead us to severely underestimate the cost of protection. Further, large trade in intermediates can lead countries to overestimate exports as a source of demand growth, but also to overlook the crucial role that imports play in enhancing efficiency and exports. Generally, the existence of large and growing trade in intermediates, which is associated with foreign direct investment and the globalisation of production, greatly raises the stakes on countries having an open and predictable trade regime.
Large trade in intermediates also has its dangers, as evidenced by the huge global trade shock imparted during the financial crisis. The answer, however, is not less trade, but building better safeguards against financial instability.
Baldwin, Richard (ed.) (2009), The Great Trade Collapse: Causes, Consequences and Prospects, A VoxEU.org Publication, 27 November.
Cappariello, R and R Zizza (2007), “Vertical specialisation in Europe: Evidence from the import content of exports”, Bank of Italy.
CBO(2008), “The domestic value added of Chinese exports”
ECB (2005), “Competitiveness and the Export Performance of the Euro Area”, ECB Occasional Paper 3
Hopenhayn, P. and A. Neumeyer (2002) , Economic Growth in Latin America and the Caribbean; Country Study for Argentina. The Role of Capital and Labor Reallocation in the Argentine Great Depression of the 1980s, Universidad Torcuato Di Tella
Maurer, Andreas and Christophe Degain (2010), “Globalisation and trade flows: what you see is not what you get”, WTO staff working paper ERSD-2010-12
Nordås, Hildegunn Kyvik (2003), “Fragmented Production: Regionalisation of Trade?”,
O’Rourke, Kevin (2009). “Collapsing trade in a Barbie world”, Irish Economy blog.
Yi, Kei-Mu (2009), “The Collapse of Global trade: The Role of Vertical Specialization” in Baldwin and Evenett (eds.), The Collapse of Global Trade, Murky Protectionism, and the Crisis: Recommendations for the G20.
1 Unless otherwise indicated, data on import content of exports is from OECD Input-Output database. It should be noted that estimates vary widely depending on the data, assumptions, and methodology used. For example, estimates of the share of Chinese exports accounted for by imported intermediates can vary from 45% to 65 %. See CBO(2008).
2 This estimate is based on five European economies – Germany, Italy, Netherlands, Austria, and Finland – which account for around 60% of Eurozone GDP. See ECB (2005).
3 The Effective Rate of Protection is given by (V* - V)/V, where V* is the domestic value added with a tariff on imports and V is the domestic value added under free trade.
4 In addition to the direct impact of higher costs on intermediate imports, which are needed for domestic firms to compete internationally, import barriers on such imports have an indirect effect on real wages of workers induced by the increase in the cost of capital. See Hopenhayn and Neumeyer (2002).