by Chakravarthi Raghavan

Geneva, 6 Mar 2000 -- The thesis that foreign investment is always good for development, and that a liberal policy towards foreign investment and Transnational Corporations (TNCs) is sufficient to achieve positive effects is challenged in a just published study by two Latin American economists.

The study shows that there is "crowding in" effect of FDI on domestic investments only where developing country-governments have maintained restrictive regimes and subjected FDI applications to screening and grant of different incentives to different firms.

But in countries that maintained or moved to 'open' regimes for TNCs and their investments, there is strong evidence of a "crowding out" effect on domestic investment, the study brings out.

The term 'crowding in' (CI) is used when the presence of the foreign direct investment by a TNC stimulates new downstream and upstream investments that would not have taken place in their absence.

A 'crowding out' (CO) effect takes place when the TNCs and their foreign investment displace domestic producers or pre-empt their investment opportunities.

A neutral effect is said to take place when a dollar of FDI leads to an investment of just one dollar in the economy.

This is an important issue in development economics and literature: if investment is a key variable in determining economic growth, does the presence of TNCs in an economy and the FDI flows associated with them result in increased total investment in an economy or reduces total investment in the economy.

The study by two Latin American economists, Manuel Agosin and Ricardo Mayer, in showing either a crowding out effect or at best a neutral effect thus challenges a premise of the push for liberalisation of FDI and policies towards TNCs by developing countries, and creating multilateral or regional rules and disciplines (on host countries).

The policies of FDI liberalisation by developing countries has been pushed from the early 1980s by the Bretton Woods Institutions (through structural adjustment and promotion of the 'Washington Consensus polices), and by the industrialized countries at the World Trade Organization.

The investment division of UNCTAD, which produces the annual World Investment Reports, has also been promoting such policies, both through the annual WIRs and other publications, and seminars and symposia.

After the collapse of the moves at the OECD for a Multilateral Agreement on Investment (MAI), to provide rights for foreign investors (including the right of entry and exit in any economy, and disciplines on the governments of host countries), the European Union and Japan have been trying to achieve the same by bringing on the agenda of the WTO the so-called "Trade and Investment" issues as well as funding related projects in UNCTAD.

But the study, 'Foreign Investments in Developing Countries: Does it Crowd in Domestic Investment', by Manuel Agosin and Ricardo Mayer (of the Department of Economics at the University of Chile, Santiago), challenges one of the central thesis of the neo- liberal economics, namely that FDI and TNC presence in an economy, and liberalisation policies to bring this about, has a positive effect on the host.

The study by Agosin and Mayer of some 32 countries in Africa, Asia and Latin American and Caribbean, published in UNCTAD's Discussion Papers series, contradicts one of the conclusions of UNCTAD's World Investment Report 1999, which discussed the same issue (pp 171-174) and cited an annexed model (189-191).

The WIR does not indicate the authors of the econometric exercise cited in support (in an annex pp 189-191), but the econometric model and exercise in the annex appears to be similar, if not the same, as in the discussion paper of Agosin and Mayer which uses a theoretical model of investment with FDI as one of the independent variable.

The somewhat different conclusion was reached by the WIR by including, in an ad hoc and unexplained way, seven more countries. The seven included are Cyprus Turkey (shown here as in Europe), Poland (a transition economy) and Oman, Saudi Arabia, Jordan and Egypt -- with all of them shown as in West Asia.

In main annex tables though the WIR classifies, Cyprus and Turkey as part of West Asia, while Egypt is classified as North Africa.

For the box, and the annex, Cyprus and Turkey are described by WIR-99, as 'developing countries' from Europe and Oman, Saudi Arabia, Jordan as from West Asia, but with Egypt thrown in this group. But elsewhere in the main annexed tables of the WIR (for e.g. Annex B at p 491), Cyprus and Turkey, with Saudi Arabia, Oman and Jordan are classified as part of West Asia, while Egypt is shown in North Africa.

This enables the WIR (in the annex of econometric analysis) to suggest in the table that in West Asia there is both crowding in and neutral effect.

The WIR draws the conclusion by dividing the effects over two periods, that Africa showed strong CI effect in the first period (1976-1985) and a weak CI effect, close to neutral in the second.

And the WIR reaches a different view (from Agosin and Mayer) by what may at best be described counter-factual based on assumptions such as 'gains in efficiency, if crowded out (domestic enterprises) are inefficient.

Agosin and Mayer note that FDI is prized by developing countries for the bundle of assets that TNCs deploy with their investments. Most of these TNC 'assets' are intangible in nature -- technology, management skills, channels for marketing products internationally, product design, quality characteristics, brand names etc -- and particularly scarce in developing countries.

But in evaluating the impact of FDI on development, a key question is whether TNCs crowd in domestic investments or they have the opposite effect of displacing domestic producers or pre- empting their investment opportunities.

This, Agosin and Mayer point, is a rather important issue, since theoretical and empirical work identify investment as a key variable of determining economic growth.

"Thus, if FDI crowds out domestic investment or fails to contribute to capital formation, there would be good reasons to question its benefits for recipient developing countries."

Given the scarcity of domestic entrepreneurship and need to nurture existing entrepreneurial talent, a finding that TNCs displace domestic firms, say Agosin and Mayer, "would also cast doubts on the favourable development effects of FDI."

These are all the more important questions, when one considers FDI is far from being a marginal magnitude, but as a share of total gross fixed capital formation is a significant and growing magnitude in the developing world, and is a much larger proportion of investment in developing than in developed countries.

FDI is a financial balance-of-payments concept, while investment is a real national accounts variable. "Much FDI," the two authors add, "never becomes investment in the real sense: mergers and acquisitions (M&As) are mere transfers of ownership of existing assets from domestic to foreign firms."

The assessment of effects of FDI on domestic and total investment is not a trivial matter, but little could be said on an a priori basis, and may vary from country to country - "depending on domestic policy, the kinds of FDI that a country receives and the strength of domestic enterprises."

But it is possible to specify conditions favourable to a CI, say the two authors. In developing country setting, foreign investments that introduce goods and services new to the domestic economy, be they for export or domestic markets, are more likely to have favourable effects on capital formation than foreign investments in areas where there are already domestic producers.

FDI could contribute to development if it introduces new goods to the economy - and with it technologies and human capital - that do not have the knowhow to human resources to produce them. But if FDI enters the economy in sectors where there are competing domestic firms, or firms already producing for export markets, the very act of foreign investment may take away investment opportunities open to domestic entrepreneurs prior to the foreign investments.

Such FDI is likely tor educe domestic investments that would have been undertaken, if not immediately at least in the future by domestic producers. The contribution of such FDI to total capital formation is likely to be less than FDI itself.

Thus, the relationship between FDI and domestic investment is likely to be complimentary when investment is in an undeveloped sector of the economy (owing to technological factors or lack of knowledge of foreign markets). But the FDI is more likely to substitute for domestic investment when it takes place in sectors where there are plenty of domestic firms or when domestic firms already have access to technology that the TNCs bring into the country.

Even where FDI does not displace domestic investment, foreign investments may not stimulate new downstream or upstream production and therefore might fail to exert CI effects on domestic investment.

"Thus the existence of backward or forward linkages from the establishment of foreign investors is a key consideration for determining the total impact of FDI on capital formation," say the two authors.

The linkages are a necessary, but not a sufficient factor for CI and, in cases where foreign firms simply displace existing ones, the existence of linkages cannot prevent CO.

Again, the impact of FDI on investment is greater when it is greenfield investments rather than when it is M&As.

Some sample studies of M&A in Argentina and Chile (in early 1990s), involving privatization of telecommunications and public utilities, showed that there was post-purchase investments in modernization and rationalization of operations.

But in the case of several other acquisitions in Latin America, the acquisition of domestic firms were almost akin to portfolio investment, with the TNCs doing nothing to improve the operation of the domestic company.

"Very recently," the authors comment, "there have been a large number of cases of FDI (involving acquisitions), all with doubtful impacts on capital formation. Many of the acquired companies are not in need of modernizing, since they operate with state-of-the art technology. Nor is it likely that their purchase by a foreign company will be followed by sequential investment that the acquired firms would not have made themselves. In such cases, the act of FDI is not investment in the national account sense, and does not lead to investments later on."

Large M&As like large portfolio inflows, might have adverse macro-economic externalities. When of a size no longer considered marginal, they tend to appreciate the exchange rate and discourage investment for export markets, and indeed for the production of importables as well.

The Agosin-Mayer paper notes in this regard that some of the most successful newly industrializing economies restrict foreign ownership: in Taiwan province of china foreign equity ownership in domestic enterprises are restricted, with no single person or entity able to own more than 15% of a domestic company and foreigners as a whole not allowed to own more than 30% of a domestic company.

Korea, until the recent financial crisis, maintained similar restrictions.

In their econometric exercise, Agosin-Mayer show that over the period 1970-1996, FDI had a CI effect in three countries of Africa (Cote d'Ivoire, Ghana and Senegal), a neutral effect on 5 African countries (Gabon, Kenya, Morocco, Niger and Tunisia) and a crowding out (CO) effect on four others (Central African Republic, Nigeria, Sierra Leone and Zimbabwe).

In Asia, there was a CI effect in three cases (Korea, Pakistan and Thailand), and a neutral effect on five others (China, Indonesia, Malaysia, the Philippines and Sri Lanka). There was no country in Asia that experienced a CO.

In Latin America, there was a neutral effect on seven countries - Argentina, Brazil, Colombia, Costa Rica, Ecuador, Mexico and Peru. There was a CO effect on five others - Bolivia, Chile, the Dominican Republic, Guatemala and Jamaica.

There was no case of CI effect in the region.

The econometric exercises, Agosin and Mayer note, suggest that over a long period of time (1970-1996), CI has been strong in Asia, and CO has been the norm in Latin America. In Africa, FDI increased overall investment one-to-one. If the exercise is done for two sub-periods separately (1976-1985 and 1986-1996), the results vary only for Africa, which then appears as having CI rather than neutral-effects.

The main conclusion that emerges from this analysis is that the positive impacts of FDI on domestic investment are not assured.

In some cases, total investment may increase much less than FDI or may even fail to rise when a country experiences an increase in FDI.

"Therefore the assumption that underpins policy towards FDI in most developing countries -- that FDI is always good for a country's development and that a liberal policy towards TNCs is sufficient to ensure a positive effect -- fails to be upheld by the data."

A separate study by Agosin, cited in the paper, shows that the most far-reaching liberalisation of FDI regimes in the 1990s took place in Latin America, while the regimes in Asia remained the least liberal in the developing world.

"Several Asian countries," Agosin and Mayer note, "still practice screening of investment applications and grant differential incentives to different firms. And some types of investment have remained prohibited for most of the period under review. Nonetheless, it is in these countries that there is strongest evidence of CI.

"In Latin America, on the other hand, these practices have been eliminated in most countries. Nonetheless, liberalization does not appear to have led to CI."

Agosin and Mayer say that while they were unable to rest the types of policies that would maximise the contribution of FDI to total investment, their analysis "does suggest that there is considerable scope for active policies that discriminate in favour of foreign investments that have positive effects on total investment."

While it was beyond the scope of the paper to discuss which policies would have these effects, they note that some countries had been successful in screening policies to ensure that FDI does not displace domestic firms or that TNCs contribute new technologies or introduce new products to the country's export basket.

But, most developing countries, they not have the administrative capabilities to implement effective screening policies and their attempts to do so often wind up scaring off TNCs altogether.

But an alternative might be to adopt a fairly liberal regime, and then go after specific companies that fit in well with the process of progressing up the quality ladder.

The two authors suggest that CI in Asia may also be associated with high overall investment rates. Where investment is strong, investment by TNCs might elicit positive investment responses in the domestic economy through backward or forward linkages.

CI may also take place in countries with low domestic investment rates, such as those in Africa, where TNCs invest in sectors that domestic investors are unable to enter - because of technological or capital requirements that domestic firms cannot meet.

This view leads to the conclusion that even in Africa, countries should not bind themselves for ever by agreeing to investment rules for all time, and for all sectors, but rather keep the options open for the future - for e.g. by liberalising in sectors where they need FDI, and closing it off in others.

"Latin America," Agosin and Mayer conclude, "is the great disappointment. One reason for the CO in that region is that overall investment has been much weaker in Latin America than in Asia. It could also be that Latin American countries have been much less choosy about FDI than Asian countries, either in the sense of prior screening or attempting to attract desirable firms." (SUNS4621)

The above article first appeared in the South-North Development Monitor (SUNS) of which Chakravarthi Raghavan is the Chief Editor.

[c] 2000, SUNS - All rights reserved. May not be reproduced, reprinted or posted to any system or service without specific permission from SUNS. This limitation includes incorporation into a database, distribution via Usenet News, bulletin board systems, mailing lists, print media or broadcast. For information about reproduction or multi-user subscriptions please contact < >