1. Introduction

This paper intends to deny the assumption made on the top of this page. That is, we will see all the way of this essay, that there is economic justification for the existence of multinational firms (MNF) in the market. We will prove it by first looking at the theoretical point of view, and then move to some cases taken in the real world, that tend to prove the assumption we are making. We will also look at some non-economical reasons to become a MNF.

2. Definition

First, we must make sure we know exactly what we are talking about. The Unwin Hyman Dictionary of Business describes a multinational enterprise (MNE) as "a firm that owns production, sales and other revenues-generating assets in a number of countries". We must clearly know the difference between a MNF and the global corporation, which operates from only one country but sells its products all around the world, fully taking advantage of economies of scale.

3. The theoretical point of view

There is a lot of paper written about the international economics. In his paper called "A Simple Theory of International Trade with Multinational Corporations", Elhanan Helpman states a model to explain whether a company should or not decide to become a MNF in order to improve its profitability.

Let’s summarize first the principal objective and assumptions of the model implied. The objective of E. Helpman was to develop a model describing the conditions where a national firm could benefit from being a multinational, and to show the pattern of trade under these conditions. Despite the amount of different studies and theories about the international economics and trade, Helpman writes that no theory tries to explain the role of the MNF in this international trade pattern. He then tries to change this situation by committing a model involving the MNF.

The first assumption made by Helpman is that we are in presence of differentiated products, economies of scale and monopolistic competition. Then, he states that some input, that is, management, marketing, R & D, can benefit the firm in more than only one plant (assuming then that we are in a manufacturer sector).

These two main assumptions are followed by a series of sub-conditions to make sure the model is working properly. First, firms try to maximize profits and doing so, want to minimize, among others, the costs of location choice. It is that way because that factor rewards tend to differ from country to country. Second, he takes a large freedom by stating that there is no transport cost, tariffs in international trade or tax differences between countries. Then Helpman states that preferences are identical everywhere.

The world where we are now is a world with two factors of production, the labour (L) and a general purpose input (H). This input is the one that can be used in more than one plant, even in different countries. This is Management, Distribution and R & D.

The plant cost function, that we will not reproduce here, says "it pays to concentrate production in a single plant unless there are transportation costs or differences across location in product price". Helpman also says that the factors of production do not move across national borders. However, the "H" input, by its nature, doesn’t follow this pattern, and then can move anywhere.

The conclusion of Helpman’s theory is that, for a certain zone in a graph representing shares of two countries in the market, there is no advantage for the formation of MNF on an economical point of view. However, when we find ourselves out of this zone, we see net advantage to get installed in another country. The Helpman’s theory states that MNF will be created if it can take advantage of the differences between factors prices in two (or more) countries by transferring activities in the cheapest country.

J. Markusen and A. Venables (1999) explain more in detail, in "The theory of endowment, intra-industry and multinational trade", the theory first developed by Helpman in 1984. More precisely, we see in this paper that there are two conditions necessary for the existence of MNF. First, trade costs must be high enough relatively to the fixed cost incurred by the building of a second plant in another country. Second, we must include the concept of the division of the world endowment between countries. As we saw in the standard Heckscher-Olin model, each country will become a net exporter of the output that uses the relatively abundant factor. However, as we also saw in the model, with a free trade, the prices will tend to equalise within all the countries. This assumption assumes that there is no transport of tariff costs. Where Markusen and Venables differ from the H-O Model is when they include the intra-industry trade factor, that is, the Headquarters’ services fees. If we link this factor to the "H" input from Helpman, authors explain the reason why small countries (they give Sweden, Switzerland and Netherlands as examples) become "overweighed" in the housing of headquarters. That is, they don’t have a lot of labour to produce, so they use this labour as HQ’s workers, using their brains to install in bigger countries the production plants. As the authors include transport cost to their model, they conclude that this is also one cause of the existence of MNF.

These transport costs create international differences in the value of the sector price index, and replace "traditional" trade with "intra-industry" trade. This intra-firm trade, as we can also call it, is of great importance in the explanation of the MNF number’s growth. That is, applying the simplest theories of trade to their own profit, firms otherwise "national" install plants in the countries where it’s more efficient, according to the H-S model, to install it!

In their paper "Financial aspects of the multinational firm, a synthesis", Khambata and Reeds explain the ATT model (Agency cost / tax shield trade-off model). From the equation [ V_L = V_U + TS – EB – AC where V_T = value of a leveraged firm, V_L = value of a unleveraded firm, TS = present value of the debt tax shield, EB = expected bankruptcy costs and AC = agency costs of debt and equity], two components are of interest for our subject. These components are TS and EB. The TS is saying, as developed by different models, that "MNFs have a greater potential tax advantage of debt relative to purely domestic firms based on the ability to borrow in multiple markets". The EB is saying, now, that "international diversification leads to a lower volatility of earnings as the MNF has cash flows in imperfectly correlated markets". This is true if a firm from a developing country diverts in a developed country, a firm from a developed country diverts to another developed country, but the situation where a firm from a developed country diverts in a developing country gives the opposite result, for obvious reason. That is, a firm with less risk gets a lower cost of capital.

Finally, the paper states different other papers in favour of MNFs economic worth. The authors quote the internationalization hypothesis ("firms choose to become MNFs rather than sell their technologies because it may be hard to price, the market may be incomplete, and/or the firm does not want to create possible future competition".), the flexibility hypothesis ("Some knowledge is tacit and non-tradable in markets and that the knowledge a firm acquires provides valuable growth options. The first investment in a foreign market provides a large within-country option for future expansion and greater operational flexibility to exploit market conditions"), the tax arbitrage hypothesis ( MNFs have the ability to shift income to minimize taxes and to circumvent market barriers) and finally the diversification hypothesis (cost of capital is lower).

As we can see, the theory is full of models, statements, assumptions, etc. saying the reasons why a national firm wants to become a MNF. Some of them, like the tax arbitrage hypothesis, are purely a selfish reason, but others rely to real worthy economical reasons, like internationalization for instance. So, the theory tends to supports the statement about the worth of MNFs.

5. The practical point of view

Now that we have seen in the precedent section that theorists find more than one economic value to implement MNFs, how all of this takes place in the real world? In this section, we will first look at what some literature says on a practical point of view, and then continue with some example from the real world.

In their paper "Multinationality and firm performance" Han, Lee and Suk empirically analyze the effects of being a MNF on the performance of firms. They introduce their paper by telling what we already know, there are two reasons why firms pursue multinationality: they "choose the lowest cost location for their operations", and they "seek to increase revenues by responding to local demand".

The previous empirical studies’ results are not as clear as these two reasons. They show mixed results, maybe because most of the empirical studies concentrated only on the United States firms, a particular country. The authors resolve this situation by using a broader sample, including the G7 countries (Canada 92 firms, France 163 firms, Germany 173 firms, Italy 60 firms, Japan 870 firms, United Kingdom 428 firms and United States 857 firms) in a single year, 1994. This single year limits the power of the results, but some conclusions can be made anyway.

They look at different elements, and how these elements were linked with the ROE (Return On Equity), a measure of performance. Using the regression approach, they tried to see how a MNF’s ROE behaves relatively to a national firm’s one. The "firm size effect" (Banz 1981) suggests that a MNF would be less profitable than a national firm (also read here a smaller firm) because its size make it difficult to show an impressive performance. That is, they control the assets size in function of the ROE given. They also control the risk of firms, with their beta factor, with an hypothesis of positive relationship between beta and ROE. According to the "agency theory" (Jensen and Meckling (1976)), a firm with more leverage should be more profitable. They analyse the debt factor over the ROE. The growth rate (sales growth) is also studied, according to the "economic life cycle theory" (Porter, 1980). A bigger growth rate means, in the hypothesis, a better ROE. They finally control on the Research and Development (R&D) expenses, with two possibilities: a negative effect on ROE because of the cost by itself, and a positive effect because of the technological advance it gives to firms.

All these controls show low effect on ROE, suggesting that there is no difference on return between a national firm and a MNF. Asset size is significant only for Japan and USA, but on a positive way. Size gives better return, in contradiction with our hypothesis. The risk is not significant at all. Debt is significant for USA, France and Italy, but with a negative relationship. Life cycle is not significant either. Finally, R&D is negatively correlated in USA only, and non-significant at all elsewhere.

The authors develop further their analysis, by dividing the ROE into its two elements: asset turnover and net profit margin (Du Pont analysis (Ross, Westerfield and Jaffe (1996))). By regressing these two elements with the factors stated earlier, we see a better result. Then, for the asset turnover (AT), size is inversely related to AT, as predicted. Beta is not relevant, debt is relevant for France, Japan and Italy, on the opposite of our prediction. Growth is relevant for Japan and UK and R&D is negatively related in three countries. Conclusion on AT is that multinationality improves Italian firms, but nothing can be said about other countries’ firms.

The regression on net profit margin (NPM) results in the fact that size matter in the United States, the Beta is still non-relevant, debt is inversely related to NPM. Growth is relevant for Italy and USA, and R&D is not related at all. Finally, the foreign sales over total sales ratio tells that for Italy, it’s good to become a MNF.

The main conclusion of the study is that there is "some" relation between performance and multinationality. However, it is not strong or universal. We can anyway state, from it, that it is "sometime" economically profitable to go international as a multinational firm.

As we know, the economical reasons explaining the existence of MNFs are internationalization, service and transfer cost of trade. We now spend our time looking at some real world examples. We will do so by studying the Canadian case, which is quite different from the European one because of its geographical situation. Its only direct neighbor is the United States of America, a big player. So we could think that a large part of the Canadian commerce would be internal and not international or external. However, the Canadian economy is definitely dedicated to international trade, partly because of the small interior market formed by its around 31 million of citizens. Most of its exports, for instance, are going to USA, as we could guess (86,81% or 207 335 million of American dollars). This being known, why do Canadian firms decide to go international by installing foreign subsidiaries instead of just exporting to its big neighbor? Actually, a lot of companies just do it! But in the so competitive and aggressive American market, exporting means getting a very good distribution network, or a good distributor.

Internationalization

However, this strategy has a big cost. Naya, a still water bottling and distributing company in the Canadian market signed an agreement with Coca-Cola to distribute the Naya Water in the States. Everything was fine, the production volume increased significantly du to big sales in USA, but when Coca-Cola decided no to renew the contract, after a failed take-over, Naya co. went very close to bankruptcy. This would not have happened if Naya co. had decided to do business by itself in USA instead. In this case, the internationalization would have been a good reason to get installed in a foreign market.

Service and transfer cost of trade

In the real world, in opposition with the theoretical world, we don’t fix the variables. They are just given and we must deal with it. For example, in the real world, who would not admit the presence of huge transport costs between countries, or even different regions in the same country? The Canadian case is telling a lot about this situation. For instance, the Province of Quebec is exporting more goods and services to the United States (52 663 022 000 $ in 1999) than in all the other provinces of the Canadian federation (38 011 000 000 $ in 1998). The reason is quite simple: it is cheaper for Quebec firms to do business with New England states than with British Columbia province or Yukon Territory.

A Canadian firm becoming a MNF is Bombardier Ltd., from Valcourt in Quebec. A look at one of its division, Bombardier Transportation, tells us that they are located in different countries, from the whole Northern America, to seven countries in Europe. Two economical reasons explain this expansion. Bombardier acquired a high knowledge of transit systems in the sixties when the city of Montreal gave this small recreational product company the contract to develop the city’s subway wagons. After this first success, the firm started to offer its services elsewhere in the World. In front of its international success, Bombardier had no choice to install plants in its new markets. The cost of transporting a wagon is too high to make it a profitable operation. Once the first plant has been set up, others followed at the same pace of Bombardier’s geographical diversification. They tough it would be easier to enter a market with a plant inside the country than to export from a European country to another. The fact that it was better to give service to its customers directly from their country was a reason for this decision. Bombardier also shows that the "H" production factor from the theory is taking place in the real world, with the majority of its engineering work being done in Canada and used in all other plants.

Even in the sole Province of Quebec, we can see the limits of trade when transport costs are great. As Krugman and Obstfelt say, "There is also little international trade in goods with high weight-to-value ratios". We could ad that it’s really hard to get a competitive product inside Quebec with this ratio. The cement industry is representative of it, while big players can’t benefit from economies of scale because of the high cost of transporting cement in the so wide French-speaking province.

We just saw the economical reasons of becoming a MNF in the Canadian perspective. There is a lot more examples we could look at in this so large country.

A few words finally on some non-economical reasons to become a MNF. The last few years witnessed an impressive number of merger and acquisitions all around the world and with firms from different countries agreeing to continue doing business as one firm. It sounds like a great idea to increase the firm’s size, but is it an economical reason? They are actually acquiring a competitor. "The main point of direct foreign investment is to allow the formation of multinational organizations. That is, the extension of control is the essential purpose." As we saw in class, world domination is not an economical reason to become a MNF. A firm that span its activities between countries to benefit from different tax treatment is not acting economically neither. For example, firms creating an expenses-intensive subsidiary in England (high taxes) while establishing the profit generating center in Bahamas (low taxes) would of course increase its performance, but without contributing to the society’s welfare.

6. Conclusion

The study of both theoretical and real world cases allows us to conclude that there are economical reasons for a firm to decide to become a multinational firm, but all the reasons do not worth the same thing. Both theory and practice however show limits to the value of being a MNF. We therefore must be prudent and state that "there may be economical reason to be a MNF", but it is a case by case situation.