Global inflows of foreign direct investment (FDI) declined in 2003 for the third year in a row, to $560 billion (table 1). This was prompted again by a fall in FDI flows to developed countries: at $367 billion, they were 25% lower than in 2002 (table 2). Worldwide, 111 countries saw a rise in flows, and 82 a decline. The fall in flows to the United States by 53%, to $30 billion – the lowest level in the past 12 years – was particularly dramatic. FDI flows to Central and Eastern Europe (CEE) also slumped, from $31 billion to $21 billion. It was only developing countries as a group that experienced a recovery, with FDI inflows rising by 9%, to $172 billion overall. But in this group, the picture was mixed: Africa and Asia and the Pacific saw an increase, while Latin America and the Caribbean experienced a continuing decline. The group of 50 least developed countries (LDCs) continued to receive little FDI ($7 billion).

As in the past, TNCs from developed countries will drive the renewed growth of world FDI flows.

But, increasingly, TNCs from developing countries are contributing too. Their share in the global FDI flows rose from less than 6% in the mid-1980s to some 11% during the latter half of the 1990s, before falling to 7% during 2001-2003 (for an annual average of $46 billion). They now account for about one-tenth of global outward FDI stock, which stood at $859 billion after rising by 8% in 2003. Measured as a share of gross fixed capital formation, some developing countries invest more abroad than some developed ones: e.g. Singapore (36%, during 2001-2003), Chile (7%) and Malaysia (5%), compared to the United States (7%), Germany (4%) and Japan (3%) (table 5). As the economic recovery takes hold, FDI from these and other developing countries can be expected to resume growth. Is a new geography of FDI flows in the making, complementing the new geography of trade?

FDI inflows to Africa rose by 28%, to $15 billion, in 2003, but fell short of their 2001 peak of $20 billion (table 2). Thirty-six countries saw a rise in inflows, and 17 a decline. The recovery was led by investment in natural resources and a revival of cross-border M&As, including through privatizations. Morocco was the largest recipient of inflows. Overall, natural-resource-rich countries (Angola, Chad, Equatorial Guinea, Nigeria, South Africa) continued to be the principal destinations, but a large number of smaller countries shared in the recovery. FDI in services is increasing, particularly in telecommunications, electricity and retail trade. In South Africa, for instance, FDI in telecommunications and information technology has overtaken that in mining and extraction.

For the fourth year in a row, FDI flows into Latin America and the Caribbean (LAC) fell, by 3% in 2003, to $50 billion (table 2). This is the lowest annual level of inward FDI since 1995. Of 40 economies, 19 saw declining inflows. In particular, declines were registered in Brazil and Mexico, the region’s largest recipients. With privatization running out of steam, weak economic recovery in the European Union (EU) (the region’s principal source of FDI, apart from the United States) and recession or slow growth in several countries in the region in the aftermath of the Argentine crisis, LAC has been hit hard by the FDI downturn. The apparent decline of the maquila industry added to concerns that Mexico might be losing attractiveness for FDI. Several smaller economies, such as Chile and Venezuela, registered increases in 2003, the former recouping its losses of the previous year. As a result, the region’s share in developing-country inflows has returned to the levels preceding the latest FDI boom. In 2003, FDI outflows from LAC rose to $11 billion.

The year 2003 saw a mixed FDI picture for the developed countries: ten posted higher inflows and 16 lower ones. Overall, inflows declined by 25%, to $367 billion (table 2). Intra-company loans plunged and, to a lesser extent, equity flows (two of the three components of FDI flows). However, reinvested earnings rose, thanks to improved profitability. The slow pace of economic recovery did not help. Crossborder M&As fell in number and value for the third year running. United States FDI inflows halved, from $63 to $30 billion, which placed that country behind Luxembourg (because of transshipped FDI), China and France. Flows into the EU as a whole declined by 21%, to $295 billion.

The structure of FDI has shifted towards services. In the early 1970s, this sector accounted for only one-quarter of the world FDI stock; in 1990 this share was less than one-half; and by 2002, it had risen to about 60% or an estimated $4 trillion (figure 3). Over the same period, the share of the primary sector in world FDI stock declined, from 9% to 6%, and that of manufacturing fell even more, from 42% to 34%.

What explains the shift of FDI towards services? Partly it reflects the ascendancy of services in economies more generally: by 2001, this sector accounted, on average, for 72% of GDP in developed countries, 52% in developing and 57% in CEE countries. Moreover, most services are not tradable – they need to be produced when and where they are consumed. Hence the principal way to bring services to foreign markets is through FDI. In addition, countries have liberalized their services FDI regimes, which has made larger inflows possible, especially in industries previously closed to foreign entry. Of particular importance has been the privatization of State-owned utilities in Latin America and the Caribbean, and in CEE.

The shift towards services is also discernible in cross-border M&As. In fact, most M&As during the second half of the 1990s took place in services and then became a widely used mode of TNC entry. While, in the late 1980s, services accounted for some 40% of global cross-border M&As, their share rose to more than 60% by the end of the 1990s. Up to the 1980s, cross-border M&As were almost exclusively the domain of United States TNCs. Since then, EU TNCs have become the dominant actors: in 2001-2003, they accounted for 61% of all M&A purchases worldwide. Cross-border M&As have also played a prominent role in the overseas expansion of services by TNCs based in developing countries.

FDI in services has traditionally been, and continues to be, market-seeking, despite the increase in the cross-border tradability of many information-intensive services. While some services (e.g. financial and, especially, business services) can be rationalized internationally, leading to efficiency-seeking FDI, the integrated production of services on the whole remains in its infancy. In 2001, for example, 84% of sales of services by foreign affiliates of United States TNCs were local sales in host countries, while the corresponding share for goods was 61%.

To start with, FDI in services, like FDI in other sectors, injects financial resources into a host economy. To the extent that funds are raised internationally, they are a net addition to resource flows into a host country. If funds are raised locally, domestic interest rates may rise, making capital more expensive for domestic enterprises, although the difference between locally-raised and foreign-sourced resources becomes less important as countries open up to international capital markets. A large part of services FDI is in market-seeking, non-tradable activities, which do not contribute directly to foreign-exchange earnings. At the same time, they entail external payments, for example, in the form of repatriated profits. Hence, FDI could have a negative impact on the balance of payments. And payments associated with FDI in services (e.g. repatriated profits) can quickly outweigh the initial capital inflow and exacerbate balance-of-payments crises.

Both direct and indirect benefits associated with services FDI can boost national and export competitiveness. However, benefits may not be realized if conditions in the host economy are not right. Services FDI can entail three kinds of risk: (i) systemic risk, when the absence of efficient regulation exposes a host economy to significant economic instability; (ii) structural risk, when the institutions and instruments needed to manage, say, privatization and utilities, are weak and there is the risk of turning State-owned monopolies into private ones; and (iii) contingent risk, when FDI in socially or culturally sensitive areas causes unintended harm.

Services typically need to be produced when and where they are consumed. In the past decade or so, advances in information and communication technologies have made it possible for more and more of these services to be produced in one location and consumed elsewhere – they have become tradable. The implication of this “tradability revolution” is that the production of entire service products (or parts thereof) can be distributed internationally – in locations offshore from firms’ home countries – in line with the comparative advantages of individual locations and the competitiveness-enhancing strategies of firms. This is a process well known in the manufacturing sector.

Cost considerations often trigger offshoring. For example, 70- 80% of companies interviewed in various studies mentioned lower costs as the main reason for setting up a shared service centre abroad. Cost savings of 20-40% are commonly reported by companies that have experience in offshoring. Savings relate both to the use of cheaper labour and the consolidation of activities in fewer locations. Hence, considerable savings can accrue from offshoring even among developed countries – where, in fact, most of it takes place.

While offshoring is creating new FDI opportunities, not all countries are taking part in this process. As with FDI and trade in general, developed countries attract a sizeable share. Given that services generally require higher skills than manufacturing activities, the barriers to entry can be high for potential host countries. For those that do manage to become export bases for services, key benefits include increased export earnings, job creation, higher wages and the upgrading of skills. Export revenues are considerable, as exemplified by India, where exports of software and IT-enabled services grew from less than $0.5 billion a decade ago to some $12 billion in 2003-2004. Jobs created in the services sector, including through offshoring, are typically better paid than in the manufacturing sector. But wage increases are also more rapid than in manufacturing, which makes offshored services more vulnerable to relocation to other sites. Given the short time needed to implement an offshore FDI project, attracting offshored services can offer fast-track job creation for successful host countries.

The growth of services offshoring has given rise to concerns mainly in developed countries. In particular, the growth of white-collar, export-oriented service jobs in some developing countries is seen as leading to employment losses in developed countries. (The benefits arising from this new international division of labour typically receive less attention.) Consequently, proposals have been made – particularly in home countries – to constrain the trend towards offshoring.

Returning to FDI in all services, there is a growing recognition by governments that, on balance, they benefit from such investment. The result has been a broad-based opening up to services FDI, although, the degree of openness varies across countries and industries. In general, developed countries are more open than developing ones. But even countries that have liberalized most of their service industries typically retain entry restrictions in specific services, such as media and air transportation. The nature of restrictions and the purpose for which these are introduced vary by industry. Services FDI can bring economic benefits, but policy-makers need to strike a balance between possible efficiency gains and other broader development objectives.

The opening up of various infrastructure services to FDI in the framework of privatization programmes has triggered unprecedented increases in such investment. While involving foreign companies in infrastructure services can bring new capital and more and better services, it can also entail costs. FDI in services through privatization raises a special challenge in terms of regulation and governance.

Over the past decade, the number of international investment agreements (IIAs) covering FDI in services has proliferated, resulting in a multilayered and multifaceted network of international rules. In many areas of services FDI, therefore, national policy-making increasingly takes place within the framework of these agreements. Agreements differ in their approach towards services FDI (investmentbased, services-based, mixed) and in their substantive provisions (e.g. regulating entry as opposed to protecting investment, adopting a positiveas opposed to a negative-list approach when making commitments). Several services IIAs contain follow-up procedures and separate chapters for specific service industries.