One of the consequences of globalization is the emergence of transnational corporations (TNCs), that operate globally with the avowed purpose of maximizing wealth for shareholders who for the most part live in Northern Cone countries. TNCs own most of the world’s intellectual property including patents, copyrights, trademarks, and trade secrets. TNCs conduct most of the private research and development efforts around the world. Bertin and Wyatt present evidence to show

The consequence of this weakness in invention in developing countries is a one way trade in technology from the developed countries to the developing countries. The one way trade results frequently in balance of payments deficits for developing countries (17, 120). TNCs claim patent rights in developing countries. Thus the IPR protection in place in the developing country helps to establish the technological dominance of the TNCs. According to Bertin and Wyatt TNCs own five out of six patent applications in developing countries. Often the TNCs purpose in filing a patent application is not to do local manufacturing with the technology; but, alternatively, its purpose is to protect an import monopoly. If the TNC does produce locally it usually exacts large royalty payments from the manufacturing subsidiary or licensee repatriating the payments to its headquarters (17, 122). If the developing country sees the situation as unfair the TNC reminds the parties that the terms of trade are dictated by the market. Thus developing countries object to the expense of enforcing IPRs when enforcement does not appear to benefit their citizens directly. TNCs complain that they are losing profits because lack of enforcement has led to the unlicensed use of their patents and copyrights. As Konan et al. tell us:

Many have noted that intellectual property has the characteristics of a public good.

Machlup wrote, "the right to exclude others from the use of particular material things is necessary for their efficient use...The rights to exclude others from the use of an idea demands a justification on alltogether [sic] different grounds (19,46). In 1962 the economist Arrow concluded that there were no grounds to exclude others from the use of ideas. Technology is information, although costly to produce. Technology is non-rival and free of opportunity cost in its use. If one allows someone else to make a copy of a blueprint or a manuscript, one is not deprived of the use of the original. Likewise the adopter does not have to pay again the costs of developing the idea. Thus, according to Arrow, technology should be free to everyone who wants it (20). Charging for the use of an idea leads to inefficient under-utilization of the idea.. If a charge is exacted existing ideas will not be diffused among firms optimally. Prospective IPR owners cannot collect royalties from all who may collaterally benefit from the technology that they create. Thus, for lack of adequate incentives, the market will supply too little intellectual property. The inventors will not consider in their decision (whether to invent or not to invent ) the various benefits that will accrue to non-owners (non-pecuniary externalities). Inventors will consider as an incentive only the economic benefits that the inventors can capture. The fact that the inventors are not able to extract all the benefits of their invention results in their choosing to produce less than the "socially" optimum amount. The under-utilization and under-production of intellectual property result from a market failure that is due to the non-pecuniary externalities that are typically associated with private production of a so-called "public" good. As Konan et al. inform us,

Some would argue that a second best solution is possible by assigning property rights over specific forms of information. When considering the debate about whether property rights should be granted in intellectual property in the form of patents, Nelson and Winter suggest that there is a tradeoff between static efficiency (holding prices at the marginal cost of production) and dynamic efficiency (maximizing gains that will result from future and continuing innovation). These authors argue that we must balance the under-utilization and lack of diffusion results in the present period against the possibility of underproduction in the future. They argue that, without the incentive patent protection offers, the system is dynamically inefficient—producing too few ideas that can be patented. Mansfield’s empirical results seem to indicate otherwise. According to his survey TNCs would continue to produce and innovate new products primarily for competitive reasons whether patent protection existed or not(22). In fact Nelson and Winters’ theoretical analysis is flawed by the implicit assumption that intellectual property is a rival good when in fact intellectual property is inherently non-rival.. Making intellectual property excludable with patent protection does not make intellectual property exactly like other goods—intellectual property retains its non-rival feature. As Konan, et al., rightly point out,

As the reader will recall Arrow argued against assigning property rights in technology. He argued that since technology is "costless to use" it should be free(20). Unfortunately, Arrow’s argument that technology should be free because it is "costless to use" ignores an important consideration. Pavitt is quick to point out this consideration,

Technology is complex because of the variety of inputs and knowledge bases required to produce it. Some of those same inputs and knowledge bases are needed to be used in the transfer the technology. Technology is tacit because of the impossibility of speaking or writing down all aspects of a technological development. Some aspects are encapsulated in the experience or know how of the inventors and the early adopters. The idiosyncratic aspect of most technological ideas mean that they cannot be communicated by simple copying but must be adapted to the situation in which they are used. On the basis of these points TNCs argue that they are the most efficient vehicles for the transfer of technology to developing countries. The TNC owns both the codified technology and the know-how that is encapsulated in the human capital of the firm. The TNC has mastered the complexity and the tacitness of its owned technology. The TNC can hire local experts to adapt to the idiosyncratic features of using the technology in another country. Unfortunately TNCs who make a direct investment in a developing country by setting up a subsidiary there, do not transfer technology to local firms. The more TNCs internalize technology and expand their market to capture the externalities associated with those technologies, the more barriers they erect to the imitation of technology by local firms. Imitation may be the way local firms can grow to become innovators and inventors themselves. This opportunity is lost if the local imitator is forced to cease operation because of a TNC held patent. TNCs will often demand patent protection for their owned technology before they will bring the technology into a developing country. The current TRIPs agreement requires developing countries to phase in western style protections for intellectual property. We speculate that this is being done at the insistence of the TNCs. The level of protection TNCs desire imposes enormous enforcement costs on the developing country not to mention the domestic product lost when local imitators are forced to cease their operations. As Konan et al. inform us,

Konan and her co-authors correctly point out that both producers and consumers in the intermediate and developing countries lose when firms that are good enough to imitate but not good enough to invent are shut down. They caution that producers and consumers throughout the world may lose as the constraints on imitation affect global economic growth. These authors imply that the TRIPs agreement may actually institutionalize constraints on global economic growth. We hasten to add that the TNCs probably do not object to institutionalizing constraints on global economic growth because TNCs will benefit as a result of the IPR ownership that constrains growth. TNCs as a result of their domination in IPR ownership see themselves getting ever-greater shares of whatever growth there is.

Developing countries are left with a dilemma. How do these countries gain access to technology without creating dependency relationships with the TNCs. It is well known that throughout the world "Improvements in technology have been the real force behind perpetually rising standards of living (1, 24)." It seems clear that a developing country in which technology is scarce could improve its standard of living greatly by simply adopting some of the technologies which exist in the developed countries. Adoption of existing technology by a developing country, however, is fraught with difficulties. Pack and Westfal advance three hypotheses concerning these difficulties:

These hypotheses—if demonstrated—suggest that the government in a developing country should engage in selective intervention to facilitate technology transfer while minimizing dependency relationships with the TNCs. This finding contradicts the neoclassical paradigm that suggests that the government in a developing country should encourage a market solution by adopting a neutral policy regime. The competing view, expressed in the Economist magazine, suggests that "macroeconomic stability, education, and infrastructure are the elements of a boring industrial policy that would work (5,15)." Neoclassicists warn against intervention in the market: "why not bestow some subsidies here or there, to give strategic industries a hand? Better not (5,16)." Neoclassicist’s worry that intervention will attract "rent seeking" behavior: "Political clout rather than economic sense decides who gets what (5,16)." The same critics note that many subsidy programs are failures in the sense that costs exceed benefits. Rather than make a country more competitive, subsidy programs inflict costs on the unsubsidized industries. Neoclassicists also emphasize that the world market must be seen as fair "if governments are to resist demands for intervention. Otherwise a Gresham’s law of government will rule, and bad policies in some countries will drive out good in others (5,16)." Thus the case for standardizing IPR regimes around the world. The problem is that standardizing intellectual property regimes will not correct the market failure that results from the "public good" characteristics of technology. As Konan and her co-authors pointed out earlier there is no international regulatory body that wishes to intervene to compensate for the market failure. Then why resist globalization. Let the TNCs bring their technology and their know-how let them internalize the externalities. This seems to be what Nelson and Winter are advocating:

This opinion is seconded by a number of other authors who according to Bertin and Wyatt,

...argue that, given technology has the properties or information,

and that the market for information is imperfect because sellers cannot

completely control its further use, multinational firms can ensure both

The problem is that this leads to "a high degree of industrial concentration" and it may also lead to dependency relationships with the TNCs who may pack up and leave a country for greener pastures any time. On the other hand, why shelter infant industries when more efficient and productive TNCs are willing to set up shop in the developing country. As the Korean case has shown there are benefits to sheltering one’s infant

industries until they can be competitive on world markets.

We join Pack and Westphal in rejecting the neoclassical paradigm in the case of developing countries. The neoclassical paradigm is based on the false assumption that a market for a specific technology and responsive agents exist from the outset in a developing country. As these authors point out:

Thus the market that exists in a developing country transmits price information only about current products. As Scitovsky observed market prices are:

...more useful for coordinating current production decisions...than they

what the future economic situation is expected to be (4.125).

The problem is that information about what the future economic situation will be is difficult and expensive to gather. The market does not transmit information about the price of future products that will exist after a technology is adopted. The costs of searching for a suitable existing technology, researching the potential market, contracting for the technology, and setting up a production line may exceed the potential gain for the average organization in a developing country. Pack an Westphal reason:

We agree with these authors’ diagnosis that there is a need for more coordination in the activities of firms in developing countries. What is necessary to facilitate technology transfer is a mechanism for economizing on the incident transactions costs of industrialization. Coordination also helps to internalize the various externalities incident to industrialization.

We point out that government selective intervention is perhaps not the only way to coordinate the activities of firms at the scale necessary. Several private sector solutions come to mind: 1) firms can engage in vertical and horizontal integration, 2) firms can form industry wide trade organizations, 3) a firm can joint venture or partner with one of the TNCs, 4) a TNC can form a local subsidiary buying the assets of many local production facilities.

Unfortunately, in the absence of government intervention there may not be adequate incentives to bring about these "private sector solutions" to the problem of insufficient coordination.

Vertical and horizontal integration of firms in developing countries is unlikely in the absence of a developed capital market. Developing countries may lack the institutions or the expertise to finance mergers and acquisitions. Overseas portfolio investment in domestic firms will be limited in the face of political uncertainty.

Effective trade organization, another possible solution to coordination, are difficult to finance in a developing country. Firms will only pay for information that is not freely available. The information that a trade organization gathers will only be partially excludable. If there are many small firms in the industry, how does the trade organization avoid the "free rider" problem? A government’s power to tax overcomes the free rider problem.

Joint ventures or partnerships with foreign technology holders (TNCs), the third possible solution, depends heavily on host country firms bringing local capital and special knowledge of local conditions to the party. The problem is that established TNCs may be better able to raise money in local capital markets than domestic firms may. TNCs may be able to hire local employees to gain local expertise without giving up ownership advantage. It may be that a TNC is able to deal with the idiosyncratic aspects of technology by training its own technology experts to deal with local conditions rather than training transferring know how in the technology to local people. In the absence of a government imposed participation requirement, joint ventures and partnerships will be rare.

Unsubsidized direct foreign investment is the only remaining "free market" private sector solution to the coordination problem. TNCs possess a ready-made hierarchy that can be extended to the Host Country. The question is what incentives does the host country offer to attract the technology owner?

Lucas (1990) points out that cheap labor may not be a sufficient incentive for direct foreign investment in the face of existing impediments. Capital flows we observe from developed countries to undeveloped countries fall far short of what we would expect on the basis of expected differences in the marginal productivity of capital. If the standard neoclassical model were a true picture of the world, enormous rate differentials would exist between the marginal productivity of capital in developing countries versus the marginal productivity of capital in developed countries. In Luca’s example he compares the supposed rate in the India versus the rate in the U.S.:

Suppose production in both these countries obeys a Cobb-Douglas

Constant returns technology with a common intercept; y =Ax^b where

Y is income per worker and xis capital per worker. Then the marginal

Productivity of capital is r = Abx^b-1 , in terms of capital per worker, and

Thus: r = bA^1/b Y^b-1/b in terms of production per worker (6,92).

If b=.4 for both countries, this "implies that the marginal product of capital in India must be about (15)^1.5 = 58 times the marginal product of capital in the United States (6,92)." With a free and complete capital market, investment goods should flow rapidly from the U.S. to India and other poor countries. Lucas points out that, if the model is correct, investment in the U.S. should stop entirely. Obviously something is wrong with the assumptions on technology and trade conditions:

... but what exactly is wrong with them, and what assumptions

Lucas considers three possible explanations. He presents evidence against differences in human capital and external benefits of human capital. He supports the notion that capital flows are limited because of capital market imperfections—most notably political risk: "only insofar as political risk is an important factor in limiting capital flows can we expect transfers of capital to speed the international equalization of factor prices (6, 96)." Thus wages in a developing country will approach the wage in the developed country and living standards will also improve if direct foreign investment takes place. Material direct foreign investment can only take place if the existing impediments to direct foreign investment are relaxed. Government policy changes may be called for to create adequate incentives to encourage (or discourage) direct foreign investments. Recent research suggests that direct foreign investment is very sensitive to government policy. Selective intervention could be used to promote the desired form of industrial organization.

Konan (1994) suggests that:

The above author sets out three possible industrial organizational arrangements for the foreign technology owner: integration by a transnational enterprise, domestic vertical integration, and imports of the intermediate good from a host country monopolist.

In a game theoretic context Konan shows that minor changes in the government policy of the host country may provide incentives for the potential donor organization to switch production regimes. "For example ceteris paribus, marginal increases in the transaction cost to foreign equity ownership from a foreign tax on direct foreign investment may result in a shift from multinational enterprise production to foreign- monopoly exports of the intermediate good or even domestically-integrated production(6,4)." Since host country monopoly exports may be the desired outcome, it seems clear that a "neutral policy" regime for the host country is suboptimal.

Here we investigate the institutional arrangements that government could promote to facilitate technology transfer to the developing country. We will review theoretical findings that bear on the question. For example we consider the effect of the structure of the economic system on decision making. We analyze the effect of the structure of agents on transaction costs. We also explore further the implications of Romer’s observation that technology is a non-rival good.

One possible approach to our investigation would be to build formal models which "cast industrial innovation as the engine of growth (1,24)." Grossman and Helpman (1994), Romer (1990), Aghion and Howitt (1992), have each developed formal models with this feature. According to Grossman and Helpman using these models:

...one can now investigate whether a decentralized market economy

This is not the course that we shall take. Instead our modest goal is to show simply how different institutional arrangements alter incentives thereby facilitating or impeding technology transfer to developing countries. We also hope to show through our arguments that a new intellectual property regime is needed to avoid constraining worldwide economic growth in the next century.

As Romer points out,

Sah and Stiglitz (1985) see these institutional arrangements as the architecture of

the economic system or organization. These arrangements:

...describe how the constituent decision making units are arranged together

The above authors refer to two specific architectures polyarchy and hierarchy.

A polyarchy is a system or organization in which there are competing decision-makers "who can undertake projects (or ideas) independently of one another (3, 716)." A hierarchy is a system or organization in which a few decision-makers undertake projects

while "others provide support in decision making (3, 716)."

Thus Romer’s "pure market"—made up of individuals—constitutes a polyarchy.

A command economy would constitute a hierarchy. The capitalist market economy is seen to have features of both polyarchy and hierarchy. In a market economy corporations are hierarchies where the actions of individuals are coordinated to a common end. These corporations, however, may have little or no incentive to coordinate their actions. Corporations compete with other corporations to achieve diverse objectives within the market economy. In this aspect a market economy resembles polyarchy. In another aspect a market system may function as a hierarchy.

Suppose the government of a market economy detects a "market failure" that results in conditions it sees as not being socially optimum, the government may use its powers of coercion to coordinate the actions of various corporations to bring about the desired result. In this instance the market economy functions a hierarchy.

Sah and Stiglitz focus on the question of how the architecture of an economic system affects the quality of decision making:

How the organizations are arranged within their respective economic systems will affect the quality of the decisions made and the nature of the errors that are made. In polyarchy, for example, if one firm fails to adopt a profitable technology, another firm within the polyarchy may accept it. On the other hand, in a hierarchical structure, "if a single bureau rejects the idea, then the idea must remain unused (3, 716)."

Sah and Stiglitz conclude that since a polyarchy accepts more good as well as bad projects compared to a hierarchy, Type II error is relatively more likely in polyarchy. On the other hand, Type I error is relatively more likely in hierarchy. Thus when the priority is to avoid Type I error, a polyarchy is preferred to a hierarchy. When the priority is to avoid Type II error, a hierarchy is preferred to a polyarchy. As in statistical testing, it is difficult to know whether to emphasize avoiding Type I or Type II error without knowing something about the loss function the host country faces.

In the case of a single technology transfer project, does the expected opportunity cost of incorrectly rejecting the technology transfer project (Type I error) outweigh the expected cost of incorrectly accepting the transfer project (Type II error)? We face the same dilemma that we face in statistical testing. Other things being equal, if we change our institutional arrangements to lessen the probability of a Type II error we increase the probability of a Type I error and vice versa. If we know our loss function however, we can estimate the respective probabilities of Type II and Type I errors that will minimize our potential loss. We can then adjust our institutional arrangements to realize the desired set of probabilities.

This intuition becomes evident if we examine a benchmark case. Suppose a developing country has a risk neutral linear utility function. Suppose the probability of making a Type I error is equal to the probability of making a Type II error. Let us assume that the developing country has scarce resources available for investment in existing technologies. Let us also assume that many other technologies with similar potential benefits exist than can be adopted with a given time period. Given these stylized facts, it appears that the expected utility loss for a Type II error will be greater than the expected utility loss from a Type I error. The opportunity cost of a Type I error is mitigated by the acceptance of an alternate technology which is equally likely to prove successful. On the other hand, if the host country makes a Type II error, it will lose scarce investment capital.

Thus if a host country finds that it is accepting too many technologies that prove to be inappropriate (Type II), it should adjust its institutional arrangements in the direction of hierarchy. If a host country finds that it is failing to adopt the desired number of existing technologies, it should adjust its institutional arrangements in the direction of polyarchy. Unless one believes that the market will perform this adjustment in institutional arrangements automatically, selective intervention by government seems to be called for.

The question remains how should government decide between institutional arrangements that have equivalent decision making characteristics? Should we move toward hierarchy by encouraging corporate mergers or should we create a public industry. Intuition indicates that government should choose the institutional arrangement that minimizes transaction costs.

In deciding whether to adopt an existing technology the host country organization must incur costs of acquiring and communicating information. These costs are characterized as transaction costs. The host country organization must determine if the technology is appropriate. A technology is deemed appropriate if 1) it is not so sophisticated that it is beyond the skill level of the available local work force 2) its required factor intensity is a reasonable match to local resources 3) local conditions permit a profitable implementation of the technology. The host country organization must determine the value of the technology. It must contract respectively with the sources of financing and the technology vendor. Other things being equal, as we said the optimum institutional arrangement can be seen to be the one that economizes these transaction and contracting costs.

Thus theory dictates two selection criteria for institutional arrangements. The first criteria is to choose from the set of institutional arrangements that render the optimal decision making characteristic. Under our stylized facts this criteria has a definite bias toward hierarchy. The second criteria is to minimize transaction costs. We can now apply these criteria to evaluate various models of technology transfer.

The first step is to identify the various modes that have been observed to transfer technology from a donor country to a host country: 1) A private firm within a host country decides to purchase or license existing technology from a firm in a donor country. This is characterized as private innovation. 2) A firm in the host country invites a firm in the donor country to partner or joint venture an implementation of a desired technology. 3) A firm within the donor country decides to set up a factory in the host country. This is characterized as direct foreign investment. 4) If no firm decides to develop a given technology within a host country, the government in the host country may set up a public enterprise to do so. This is called public innovation. 5) If the government does not want to undertake the project itself, the government will provide incentives to coordinate the actions of private corporations. This we have characterized as selective government intervention. Having identified candidate institutional arrangements, we can proceed to compare these arrangements with respect to decision-making effectiveness and with respect to their performance in economizing transaction costs. Firs let us identify the decision-making organizations and the transaction costs involved.

Each mode requires at least one organization to make a decision. Each mode has associated transaction costs. Some modes or transfer may entail significant externalities that are not internalized.

Acquiring existing technology is not a passive, costless, riskless endeavor.

In the first place, the host organization’s decision to adopt a technology must be made with only imperfect knowledge. As Arrow points out adopters "cannot know the true value or the information until they possess it (4, 121)." In the second place, the host organization contracts with the technology owner in the face or all the problems that are associated with asymmetric information. As in George Akerloff’s "market for lemons" the seller of a technology knows more about the value and utility of that technology than the prospective buyer does. In the third place, communication between individuals and firms is costly and imperfect. In the fourth place, polyarchy and weakly coordinated hierarchies present the possibility of redundant adoptions of a technology with a consequent loss in allocative efficiency. In the fifth place the incentives to innovate may not be strong enough to justify the initial outlays.

Sah and Stiglitz point out that information costs include direct costs measured in time and resources and indirect costs measured as the opportunity cost of information lost due to imperfect communication. A separate issue is the bounded rationality of

Individuals. There are limits to the information absorption and processing capabilities of individuals. Bounded rationality explains why an organization of individuals may be able to gather and process more information than an individual—thus possibly making better decisions than an individual could. These authors are careful to point out that because communication is costly and imperfect:

The balance of Sah and Stiglitz’s analysis of institutional arrangements abstracts away from externalities and nonrival goods—stylized facts that we deem important to the case of technology transfer. We can, however, apply the principles introduced thus far to a heuristic evaluation of the institutional arrangements we have enumerated above.

Private innovation in developing countries is expected to perform relatively poorly in a decision making sense. This intuition depends on the stylized fact that firm sizes in developing countries are too small to "appropriate all the benefits of investments in information gathering." Other firms may free ride on the host organization’s discovery of an appropriate technology. Scarce capital prevents the firm from exploiting all the useful information it uncovers. In addition, "to the extent that information search has substantial fixed costs, individual firms cannot support the volume of information search needed to realize scale and scope economies. Thus there is inadequate coordination of the information search process. We thus characterize the economic system as more similar to polyarchy than hierarchy. Polyarchy is suboptimal in that it allows a higher incidence of Type II-errors. Without government intervention, too many inappropriate technologies will be adopted wasting scarce capital. Although attempts to transfer technology may be frequent, successful transfers of technology are expected to be rare. This will not be the case in developed countries where individual firms are large enough to support adequate coordination in information search within a single firm.

Private innovation in developing countries is also expected to perform poorly at economizing on transaction costs. The cost of searching for an appropriate technology is high. The local advantage of domestic firms is often offset by high cost of acquiring existing technology. Pack and Westphal observe that:

Firms must gain the technological capability to assimilate existing technology. These authors emphasize that "not all knowledge can be put in communicable form (4, 108)." This feature is referred to as the tacitness of technology. "The tacitness of important elements of technology is in fact the principal reason for dynamic economies in learning to use ‘known’ (somewhere else in the world) technologies (4, 108)." The acquisition of technology is a sequential process; thus, the experience, which is gained over time, accumulates creating "dynamic economies." Dynamic economies also result from use of existing technology to produce new elements of technology.

Because of the tacitness of knowledge some aspects of technology are either nontradable or imperfectly tradable. For example knowledge of how to adapt technology to local resources or circumstances is imperfectly tradable. "The fact that technological elements are not perfectly tradable means that investments to acquire and assimilate them can generate surpluses for those undertaking the investments or for other beneficiaries (4, 109)." These surpluses constitute producer’s surplus and user’s surplus respectively. These surpluses may impart a competitive advantage over imports either in price or quality. In developing countries firms are often too small initially to capture significant surpluses. Since these surpluses are the result of dynamic economies, government intervention to protect an infant industry could allow these surpluses to accrue until the industry became internationally competitive.

Without government intervention the amount of private innovation undertaken by firms is suboptimal. The reason is that many of the benefits to innovation accrue outside the firm in the form of non-pecuniary and pecuniary externalities. Non-pecuniary externalities are externalities that occur without a market transaction. For example, company A’s investment in technological information reduces transactions costs for companies B through Z wishing to acquire similar information. Pecuniary externalities are the result of market transactions such as the sale of elements of technology or technological services. Such a sale may be made either within the same industry (intra-industry) or to another industry (inter-industry). Technological elements may also be transferred in the sale or exchange of intermediate and capital goods. These externalities constitute an inability of the firm to appropriate all of the benefits that result from its activities.

Thus it appears that without government coordination private innovation in developing countries will fail to provide the socially optimal amount of investment in technology transfer. A partial solution is for government to coordinate the actions of firms or to encourage mergers. If government assists firms to become larger and more hierarchical, firms will be greater able to economize on transactions costs. Larger firms will be able to acquire and assimilate more information. These firms will have a better chance of becoming internationally competitive.

We commented earlier that donor country organizations would be unlikely to seek local partners unless encouraged by the government of the host country. If so encouraged, the resulting partnerships or joint ventures should perform well. Partnerships or joint ventures as a method should work better than private innovation. The Hybrid Company has the advantage in decision making. The hybrid should be better able to economize on transactions costs and to capture and internalize externalities.

The hybrid has the advantage in decision making, assuming it is more hierarchical. Hierarchies as you recall are preferred to polyarchies where the goal is to avoid choosing to accept inappropriate technologies (Type-II error). In addition, the hybrid undoubtedly employs experts at the top who through experience are able to assimilate and process more information than the neophytes at private firms in the host country. The Hybrid Company also has local expertise. Perhaps local personnel are sent to work at factories or labs at technology centers within the donor country. As a result, these local personnel are better able to adapt the technology to local conditions having observed the technology in operation. Because these personnel work for the company, few significant details of the process or improvements in the process will be withheld from them.

A partnership arrangement is also superior to private innovation in economizing on transactions costs in capturing potential externalities. Some of the tacitness of

technology can be overcome by importing key personnel from the donor country.

This economizes on the transactions costs of communicating the technology. In addition, once the terms of the partnership or joint venture are set, further contracting costs are minimized. Since a joint venture or partnership represents a long-term commitment with closer performance monitoring, the Donor Company is more likely to consider moving high technology stages of its production process to the host country. It is also more likely to investigate how to adapt more advanced technology to local conditions, skill levels, and factor intensities.

As an additional benefit, transfer pricing in the export of intermediate or final goods to the donor firm will in some cases allow the donor firm to benefit from a lower tax rate in the host country or vice versa. By setting the transfer price high for intermediate good shipped to the donor country, the hybrid retains more of its income in the low rate country to assist in future growth. Government can encourage partnerships or joint ventures by offering tax concessions or tax holidays to the donor country organization.

Since the hybrid firm is more likely to be able to seek a global market, it is better able to appropriate the benefits of an increasing returns to scale technology if it exists. There will be fewer uncaptured pecuniary and non-pecuniary externalities compared to the case of private innovation. On the downside a hybrid organization may experience costs due to a clash in the corporate cultures of the two organizations. Local managers may resist the layoff of redundant employees for example.

Although the partnership or joint venture seems a good choice, some donor corporations will not invest in hybrid enterprises. These donor country firms do not want to lose control of their technology. These donor firms prefer to retain an ownership advantage by making a direct foreign investment in setting up a subsidiary in the host country. Such firms realize that it is often easier for them to raise money in host country capital markets than it is for a competing host country organization. Donor country firms also have an advantage over local firms in that they have full access to world equity and debt markets. Host country firms may not be well established enough to have full access to world equity and debt markets. Even if significant donor portfolio investment will increase the total level of investment in technology transfer to the host country. Thus a host country may wish to encourage direct foreign investment.

Direct foreign investment has a decision making advantage over private innovation. It may not be superior in decision-making performance to a hybrid organization. The advantage over private innovation is that the donor country organization is more hierarchical than the local is firm and thus less likely to commit Type-II error by transferring an inappropriate technology. The possible disadvantage in decision making compared to the hybrid organization is due to an initial lack of institutionalized knowledge of local conditions. This handicap can be partially overcome by hiring local experts. To do this however the firm must invest in searching for and contracting with these experts. The experts the donor hires may not be the best experts available. To furnish additional incentives for direct foreign investment, the government could furnish pre-screened and tested local experts to donor organizations. These experts would help to facilitate the choice of appropriate technologies. The experts would assist in the dynamic adoption process. These government provided experts could eliminate the decision making disadvantage the donor firm suffers in comparison to the hybrid firm. Government provided experts could also level the playing field in the transaction and contracting cost arena.

Direct foreign investment may economize transactions costs better than private innovation. The larger more hierarchical donor organization is set up to process information more effectively than the less hierarchical local firm is. As discussed above the technology owner can economize on the communication costs a local firm would incur in licensing technology due to the tacitness of the donor technology. It also seems clear that without government assistance the donor firm would still be at a transactions cost disadvantage to a hybrid firm other things being equal. Unpartnered donor firms must invest in a search for experts in local conditions. These firms may incur additional costs if the wrong experts are chosen. Donor firms must also invest in contracting with these experts who each may demand a royalty.

A possible disadvantage of direct foreign investment from a host social welfare standpoint is that donor firms may not have sufficient incentives to choose socially appropriate technologies. Donor technology may be capital intensive—for example. The cost of adapting the capital-intensive technology to employ more of the locally abundant factor—may outweigh the benefits to the firm. Unfortunately, intensive use of locally scarce capital by the donor may create shortages of capital for local firms. In addition to donor’s failure to use abundant labor will have adverse effects on income distribution. This will subvert the social purpose of development, which was to raise the average standard of living. In this case, fewer people may be hired by the donor firm than is socially optimum. Because capital is bid away from labor-intensive local industries, jobs may be lost in local industries. Domestic markets may be hurt by a collapse in local purchasing power. A hybrid organization is only somewhat less likely to adopt socially inappropriate technology. It may be more prone to adapt technology to local conditions. However, a hybrid firm still may not have sufficient economic incentives to adapt the technology to its socially optimum state Private innovators may also behave in a socially suboptimal way given the lack of economic incentives to choose the socially preferred outcome.

If a host government is certain that the incentives are not adequate for any firm to undertake innovation of an appropriate technology, which it sees as essential to the progress of the country, the government may adopt a strategy of public innovation.

Public innovation may have a performance advantage in decision making over the three modes previously discussed. Government hierarchies because they are more hierarchical than private firms may have an advantage at avoiding the acceptance of inappropriate technology (Type II-error). This would certainly be true in the absence of rent seeking behavior. However, socio-political factors become important here. Does the society pull together or are political corruption and rent seeking out of control. The success of public innovation like other forms of government intervention is dependent on socio-political factors. Government policies evolve through a give-and-take process among independent interest groups. Government may only undertake projects that are politically feasible. Furthermore, governments differ in their skill at implementing intervention policy. For these reasons, Pack and Westphal note that "sociopolitical factors are the primary determinants of the efficacy of different forms of government intervention (4, 103)."

Because of socio-political factors public innovation may not economize on transactions costs in comparison with hybrid innovators or direct foreign investors. Although contracting and communication costs are minimized in a public setting, the need to obtain a political consensus before establishing or expanding the venture may be very expensive in terms of the opportunity cost of production lost during the protracted decision making period. In addition, multiple interest groups may waste resources in rent seeking.

Once established a publicly operated firm may still maintain an advantage in economizing transactions costs over private innovation in developing countries. Private firms, as we said earlier, may not have the information processing capability of larger hierarchies. Private firms may not be able to take advantage of scale and scope economies in information gathering. The government is able to take advantage of scale and scope economies, however, the government may become unmanageable if it tries to do everything. For this reason selective intervention by government is preferred to public innovation.

Selective intervention by government can promote institutional arrangements which will be more appropriate for decision making than those which will exist under neutral policy regime. Selective intervention may also promote institutional arrangements that minimize transactions and contracting costs. Selective intervention may also be used to furnish incentives for firms to adopt socially appropriate technologies. Selective intervention may be used to promote partnerships and direct foreign investment. Selective intervention may be used to maximize social welfare.

Private firms in developing countries are not as able to acquire technical information as their counterparts in developed countries. As Pack and Westphal observe,

Government coordination of the technology search of private firms creates a hierarchy that is superior from a decision-making standpoint to the polyarchy, which exists without coordination. With government coordination, private industry may avoid accepting inappropriate technologies (Type II-error).

Selective intervention also allows firms to economize on transactions and contracting costs. The economies of scale and scope to be realized by coordination of the technology search effort have been previously mentioned. Government guarantees of intellectual property rights reduce the cost of contracting to license technology from donor countries. Intellectual property rights laws also provide additional incentives to make specific adaptations of technology for the local market.

Selective intervention may also correct for reciprocal pecuniary externalities. It often happens that the comparative advantage of an upstream industry depends on a downstream industry also being present in the host country. These inseparabilities arise as a result of transportation and other transactions costs as well as qualitative differences in products or technologies. Coordination is need to provide information that goes beyond the price and quantity information available in the market. Pack and Westphal note that:

Pack and Westphal also observe another reason for selective intervention in controlling direct foreign investments:

....appropriate local strategies may not be identified by foreign agents... (4, 114)

An additional advantage to developing technological capability locally is the bargaining advantage it gives in acquiring or licensing technology. The more local technological capability that exists, the less asymmetry of information between the buyer in the host country and seller in the donor country. Some technological capability is necessary to conduct an effective search for technology. More experience with technology will allow agents to assimilate more advanced technology. There are many such dynamic economies in industrialization.

Pack and Westphal point out that it is difficult to predict in advance ho a dynamic economy will be distributed between a pecuniary externality and an internal economy. Substitution of internal markets through organizational integration does not always minimize transactions costs. Large firms may invest in costly duplicate facilities and technological capability that could be shared with other firms if work was subcontracted. Selective intervention may be used to balance the potential gains from organizational integration through corporate mergers and acquisitions and the potential gains from specialization through subcontracting and consulting arrangements.

Another case for selective intervention is the observation that asymmetric information leads to second best equilibria in credit and financial markets. These equilibria are second best because of distortions in the mechanism that allocates resources. As Mattessini observes,

Loan guarantees, for example, can be used to correct a market failure that results in too few loans being made. "The reason why the market failure can be corrected in this case is that government requires a different information set than do single banks (8,72)." The bank is concerned with the probability of repayment of individuals where the social planner is only concerned that the social return on the pooled investment exceeds the opportunity cost.

Flam and Steiger (1989) and W.Gale (1989) also explore the issue of selective intervention in credit markets characterized by asymmetric information. Flam and Steiger explore the situation in developing countries where infant industries find it difficult to grow because they cannot qualify for external financing. In a manner similar to Stiglitz and Weiss (1981), these authors look at the market failure that arises from the fact that, since banks cannot observe individual firms, the local interest rate must be based on the average default probability of the pool of borrowers. Firms with low default probabilities may discover that their investment projects have a negative net present value at the prevailing loan rate even though the discounted expected return exceeds the social cost of the funds. In this sense society may be underinvested. Flam and Steiger show that a small tariff on competing imports would lower the cost of borrowing for all firms and would induce firms who would not otherwise enter the market, but who could qualify to borrow, to demand funds thereby increasing social welfare.

Gale (1989) alternatively, proposes a self-selection model where collateral is used as a screening device. Loans to high-risk firms involve no collateral while loans to low risk firms require collateral. A guarantee to low risk borrowers reduces the interest rate to them, however, it also makes this contract attractive to high-risk borrowers. To restore incentive compatibility the lender raises the collateral requirement. The use of collateral implies a deadweight loss to society--thus an increase in collateral requirements means that offering guarantees to low risk borrowers reduces welfare. On the other hand, a guarantee to high-risk borrowers causes lenders to reduce the collateral requirement for low risk borrowers by making high-risk loans more attractive to the lender and thereby relaxing the incentive compatibility constraint. Thus if equilibrium implies rationing, Gale shows that government loan guarantees will result in an increase in credit rationing.

Additional work needs to be done on the efficacy of intervention in credit markets using the costly state verification approach that treats financial intermediation as an endogenous response to asymmetric information in the credit market.

Another issue in selective intervention is the policy challenge presented by the nonrivalry and occasional nonexcludability of technology goods (11,16). Commercial goods are rival and excludable. Public goods are nonrival and nonexcludable. Technology goods are nonrival and either excludable, partially excludable or nonexcludable. Romer points out, for example, "computer software is a nonrival good, but it can be made excludable using copy protection or copyright (12.98)."

According to Romer, for growth theory it is partial excludability not partial rivalry that is relevant. While new mechanical designs, semiconductor chip designs, chemical processes, and other technological innovations are purely nonrival, unlike basic scientific discoveries, these innovations are partially excludable. Otherwise, wealth-maximizing firms would not invest in producing these inventions. He points out that Japan spends a larger proportion of their GDP on proprietary research that generates excludable benefits. Most experts agree that Japan’s growth is enhanced by institutional arrangements that promote commercial research and development in excludable technology (12,98)

The other half of the story is that the nonrival aspect of technological goods creates problems for the social policy maker. Policy makers would like to use aggregate equilibrium theories based on price taking. Unfortunately, the nonrival aspect of the technological inputs to production create a global nonconvexity in the production function which precludes the price-taking model. Romer reasons that it should be possible.

Thus the elasticity of output is greater than one and the production function is not concave. This nonconvexity cannot be explained away. It derives from the nonrival nature of the technological inputs. If excludability were the only problem, we could solve it in the same way we solve the open field problem by assigning property rights to the field. If the good was by nature a nonexcludable good such as national defense, the government would step in to provide it using its powers of coercion to tax everyone to pay for it. Thus price taking would take place in the markets for all of the excludable goods. The government would provide the public good to achieve a Pareto optimum solution. Excludability is not he only problem with technology. Since nonrival inputs are present, this changes the picture. Nonconvexities matter for aggregate level analysis. Imperfect competition is implied by increasing returns to scale technology. Imperfect competition is also implied by the market power granted the owners of patents or trade secrets. For a firm that uses nonrival technology as one of its inputs to production, selling its goods at marginal cost is not feasible since marginal cost is below average cost. An economy with imperfect competition will not exhibit price taking. Equilibria will generally not be Pareto optimal (16, 56). The nonrivalry of technological goods has important implications for government policy. Perhaps the most important point is that increasing returns implies there are great benefits to international trade. The larger market implied by trade allows the competitive firm to exploit scale economies. It also appears that trade in goods between similar countries leads to "a welfare-improving reallocation of resources used in research (12,101)." Theory suggests that a developing country should seek export led growth. Thus selective intervention should be focused on creating and maintaining international competitiveness in key industries.

Developing countries should be able to find ways to become competitive in some key industries without creating too many dependency relationships with TNCs. Granting TNCs patent rights may lower the cost of licensing technologies by reducing the IPR owners risks; however, putting local grown imitators out of business may prevent these imitators from ever making their own contribution in terms of invention. This recalls Konan and her co-authors point that a "healthy direction" for policy reform is to retain flexibility. It should be understood that the optimum length, breadth, and scope of IPR protection varies across industries, products, and countries. As the TRIPs agreement that calls for stronger IPR protection is phased in "developing countries may incur welfare losses(18,28)." Konan and her co-authors admit that there is research that suggests that,

a strengthening of IPR protection would increase trade flows...suggesting

a global improvement in allocative efficiency resulting from stronger

IPR protection. (18, 20).

However, these authors warn that "As yet there are no satisfactory conclusions regarding

IPRs and FDI [foreign direct investment] flows (18,20)."

We argue that stronger IPR protection favors the TNCs who already dominate telecommunications, media, transportation and other key sectors around the globe. Besides creating a suboptimal industry concentration in industries that are not natural monopolies, we argue that ownership of key industries by TNCs has the disadvantage that this ownership creates a dependency relationship between the developing country and the TNC. The TNC after all has no loyalty to the nation-states in which it operates. The TNCs stated goal is to maximize wealth for its shareholders. If it suddenly becomes more profitable to operate somewhere else than in the developing country, the TNC may simply sell off its assets and move its capital and expertise elsewhere. The flight of TNC capital and expertise from a country could have catastrophic effects on the lives of the indigenous populations. People who have adapted their lives to become factory workers may be forced to go back to a life of subsistence farming.

In the short run developing countries will undoubtedly lose from enforcing stronger IPR protection. The amount of money that they will have to spend on enforcement alone will be a drag on their economies. Few countries can afford to spend the $300 million dollars a year that the U.S. spends to maintain it patent system. To put a patent system in place it must also be supported by a court system. Additional law courts are another expensive proposition for developing countries. The impact of more enforcement may be merely to change the shape but not the volume of piracy efforts in developing countries. Konan and her co-authors suggest that Chinese CD pirates may simply quit using large factories and disperse production over smaller less easily detected facilities (18,33). Instead of wasting resources on misdirected enforcement efforts these authors suggest a "healthier direction would be to "focus refinement, implementation, and further changes in TRIPs on arrangements that accelerate innovation in both developed and developing countries. (18,33)."

Just what those arrangements should be is a challenge to further research. The above considerations suggest that it is time to invent an international IPR regime that will not be a drag on the world’s economy. It is time to invent an IPR regime that will not perpetuate the wealth of the Northern Cone countries at the expense of the Southern cone countries. Will the TNCs continue to merge until Pluto, Incorporated owns everything including an option on the ideas yet to be thought up. One favorable impact of Globalization is that global media have made us aware of the plight of our neighbors in developing countries. We have also become more aware that our economy and fate is linked with theirs. Perhaps if we fully appreciated the "nonconvexities" inherent in the use of technology the Northern Cone countries and the TNCs would realize that it is within our power to lift those countries out of poverty by simply sharing our technologies with them. Instead TNCs are racing to claim the last opportunity to exploit their neighbors. Sadly no one seems to realize the self-defeating nature of this quest.

References

1. Grossman, Gene M. And Elhanan Helpman, (1994), "Endogenous Innovation in the Theory of Growth," Journal of Economic Perspectives, Winter 1994, 8:1, 23-44.

2. Romer, Paul, (1993), "Implementing a National Technology Strategy with Self-Organizing Industry Investment Boards," Broodings Papers: Microeconomic 2, 1993, 345-390.

3. Sah, Raaj Kumar and Joseph E. Stiglitz, (1985). "The Architecture of Economic Systems: Hierarchies and Polyarchies," Economic Growth Center Discussion paper No. 476, Yale University.

4. Pack, H and L. E. Westphal , (1986), "Industrial Strategy and Technological Change," Journal of Development Economics 22 87-128, North-Holland

5. The Economist, June 8, 1991, "Of State and Industry," 15-16.

6. Konan, Denise Eby, (1994), "International Vertical Integration: A Positive Model of Endogenous Market Structure," typescript: The East-West Center.

7. Akerloff, G. (1970), "The Market for Lemons," Quarterly Journal of Economics 84, August, pp. 488-500.

8. Mattesini, Fabrizio, (1993), Financial Markets Asymmetric Information and Macroeconomic Equilibrium, (Dartmouth: Brookfield).

9. Flam, H. And Steiger, R. W. (1989), Adverse Selection in Credit Markets and Infant Industry Protection, NBER workin paper no. 2864.

10. Gale, W. G., (1989) Collateral, Rationing, and Government Intervention in Credit Markets, NBER working paper no. 3083, August.

11. Romer, (1994), "Endogenous Innovation in the Theory of Growth," Journal of Economic Perspectives 8: 1, Winter 1004, 23-44.

12. Romer, (1990), "Are Nonconvexities Important for Understanding Growth?" American Economic Review, 80:2, May 1990, pp97-103.

13. Collins, Susan, (1990), "Lessons From Korean Economic Growth," American Economic Review, 80:2, May 1990, pp. 104-107.

14. Park, Yung Chul, (1990), "Development Lessons from Asia: The Role of Government in South Korea and Taiwan," American Economic Review, 80:2, May 1990, pp. 104-107.

15. Agion, Philippe and Howitt, Peter, "A Model of Growth Through Creative Destruction," MIT working paper, 1989. Quoted in 12.

16. Romer, Paul M. , (1987), Growth Based on Increasing Returns Due to Specialization," American Economic Review Proceedings, May 1987, 77, 56-62

17. Gilles Bertin and Sally Wyatt, Multinationals and Industrial Property: The Control of the World’s Technology, Atlantic Highlands, NJ, 1988.

18. Denise Konan, Sumner La Croix, James Roumasset, Jeffrey Heinrich, "Intellectual Property rights in the Asian-Pacific Region: problems, patterns, and policy," Asian Pacific Economic Literature, volume 2, number 9, September, 1995,

19. Ulf Anderfelt, International Patent Leislation and Developing Countries, The Hague, 1971.

20. Kenneth Arrow and R. Nelson, "Economic Welfare and Allocation of Resources for

Invention, " in The Rate and Direction of Inventive Activity, Princeton, 1962.

21. R. Nelson and S. Winter., An Evolutionary Theory of Economic Change, Belknap, 1982.

22. E. Mansfield, "Patents and Innovation: an Empirical Study," Management Science, 32:173-81.

23. E. Mansfield and M. Schwartz and W. Wagner, "Imitation Costs and Patents: An Empirical Study, Economic Journal Vol. 91, pp. 907-18.