Public incentives in support of R+D and innovation in the Federal State of Espirito Santo, Brazil

Content

List of figures

List of tables

1 Introduction

2 Explanation of basic terms and theories
2.1 R&D and innovation
2.2 Innovation, technology and science policy
2.3 National Innovation Systems

3 Should the government support R&D and innovations?
3.1 The arguments for public support
3.1.1 Externalities
3.1.2 Public goods
3.1.3 Market structure
3.1.4 Imperfect information
3.2 The arguments against public support

4 The innovation policy tools
4.1 Financial measures
4.1.1 Public funding
4.1.2 Venture capital
4.1.3 Tax incentives
4.2 Non-financial measures
4.2.1 Public procurement
4.2.2 Human resource development
4.2.3 Laws and regulations
4.2.4 Industrial standards
4.2.5 Promotion of co-operations
4.2.6 Technology transfer
4.2.7 Technological infrastructure
4.2.8 Cooperative measures
4.2.9 Supporting policies
4.3 Final considerations

5 The federal state of Espirito Santo
5.1 General data about Espirito Santo
5.2 Economic data
5.2.1 Important peculiarities of the Brazilian economy
5.2.2 The economy of Espirito Santo
5.3 The regional innovation system of Espirito Santo
5.3.1 The weak points of the Capixabian Innovation System
5.4 Espirito Santo and innovations: the current situation

6 The innovation policy tools currently used in Espirito Santo
6.1 What Brasilia is doing
6.1.1 Tax incentives
6.1.2 Financing and financial support
6.1.3 The risk capital program
6.1.4 Scholarships
6.1.5 Innovation prizes
6.1.6 Innovation law
6.2 What Espirito Santo is doing
6.2.1 Instruments employed by IPES
6.2.2 Instruments employed by the CMCT
6.2.3 How TecVitoria is supporting R&D in Espirito Santo
6.2.4 Instruments employed by other public institutions

7 Evaluation of the incentives in support of R&D and innovation currently used in ES and suggestions for improvement
7.1 Evaluation of the financial measures
7.1.1 Evaluation of public funding
7.1.2 Venture capital
7.1.3 Evaluation of tax incentives
7.2 Evaluation of the non-financial measures
7.2.1 The evaluation of the instruments to support human resources
7.2.2 The evaluation of the laws and regulations
7.2.3 The evaluation of the industrial standards
7.2.4 The evaluation of the instruments to promote co-operations
7.2.5 The evaluation of the instruments to support technology transfer
7.2.6 Evaluation of the tools to promote an innovative infrastructure
7.2.7 Evaluation of the cooperative measures
7.2.8 Evaluation of the instruments to support SMEs

8 Summary and main conclusions

9 References

10 Appendix - Questions for the in-depth interviews

List of figures

Graphic 1: The innovation process today – innovation in the broader sense

Graphic 2: The influence of the different policies on the innovation process

Figure 1: The effects of an R&D subsidy

Figure 2: The effects of an R&D subsidy

Figure 3: The location of ES inside Brazil

Figure 4: The location of Vitória inside ES

Diagram 1: The official inflation rates since January 2002

Diagram 2: The development of the interest rate in the last months

Diagram 3: Public sector net debt in % of GDP

Diagram 4: Distribution of company size in ES

Diagram 5: Contribution of the sectors to the GDP of ES

Diagram 6: The most important actors of the Capixabian IS and how they interact

Diagram 7: The R&D expenditure of Brazil and other selected countries as % of the GNP

Diagram 8: The R&D expenditure of Brazil and other selected countries as % of GNP (1987-97)

Diagram 9: The five steps of incubation at TecVitoria

List of tables

Table 1: The instruments of innovation policy

Table 2: The R&D expenditure of ES compared to those of the other federal states (in 1000 R$)

Table 3 : The relative R&D (per R$ of the GDP) expenditures compared to those of the other federal states

Table 4: Patents granted in the USA to Brazil and other selected countries

Table 5: Patents applied for by Brazilian companies in the USA vs. in Brazil

Table 6: The density of scientists and engineers working in R&D absolute and (per 10.000 labor force)

Table 7: Numbers of articles published in international journals

Table 8: The instruments available to support R&D and innovation in Espirito Santo

1 Introduction

The purpose of this paper is to analyze and evaluate the innovation policy instruments used to promote R&D and innovation in the federal state of Espirito Santo, Brazil. The choice of the most effective instruments is of great importance for the development of the federal state as innovation is widely regarded as one of the main keys to competitiveness and economic growth of companies as well as countries. Although the promotion of R&D and innovation is essential to every economy this is even more the case for Brazil. The reason for this is simple. After decades where the Brazilian industry was protected against competition from the developed countries, the opening of the Brazilian markets in the beginning of the 90’s led to competition from foreign companies that employ the latest technologies while the Brazilian firms have to rely on technologies that still correspond to that of developed countries about 30 years ago. Therefore, R&D and innovation for the technological development of Brazil is a must. This paper will discuss if Espirito Santo is using the right incentives to achieve this.

The paper is divided into a theoretical and a practical part. The theoretical part should provide the basis for the evaluation of the innovation policy of Espirito Santo. In the practical part the information obtained at interviews with the private and public institutions influencing R&D and innovation in Espirito Santo will be used and be supplemented by data from other sources. The theoretical part starts with a description of the innovation process. After that follows the explanation of the concept of a National Innovation System. Then, I will go into the question if governments should support R&D and innovation or if the market mechanisms are able to guarantee a sufficient level of R&D. Finally, the current innovation policy instruments should be described and their advantages, disadvantages and peculiarities be explained. The theoretical part first focuses on general data about of Espirito Santo and then describes the economic conditions of Brazil relevant to Espirito Santo too. This follows a more detailed analysis of the economy of Espirito Santo. After that the institutions belonging to the RIS of Espirito Santo should be presented and their relationships should be shown graphically. Then, we will have a look at the current situation of Espirito Santo in connection with R&D and innovation. Following this the innovation policy instruments used in Espirito Santo will be described in order to evaluate how well they suite to the socioeconomic conditions of the federal state. Finally, the most important findings will be summarized and several conclusions from the evaluation of the instruments will be drawn.

2 Explanation of basic terms and theories

Before being able to discuss the various incentives respectively the different instruments that can be used in support of R&D and innovations it is necessary to define the relevant terms and explains the theories. In this context I will first explain what is meant by R&D and innovation and also how these two are interrelated. Following this I will describe how the terms research policy, technology policy and innovation policy will be used in this paper and how they are connected to each other. Finally, I will devote some attention to an important concept, namely that of National Innovation Systems.

2.1 R&D and innovation

The Organisation for Economic Co-operation and Development defines in its Frascati-Manual research and development (R&D) as "creative work undertaken on a systematic basis in order to increase the stock of knowledge...and the use of this stock of knowledge to devise new applications." (OECD, 2002, p.30). It consists of basic research, applied research and experimental development. While the objective of basic research is to acquire new knowledge without any particular application or use in view, applied research is directed towards a specific practical aim or objective. Experimental development can be seen as a systematic work based on existing findings from earlier research or practical experience which has the aim to produce new materials, products or devices, and install new or improve already existing processes, systems and services (OECD, 2002). As will be shown later, the distinction between the various components of R&D is of great importance regarding the discussion about the pros and the cons of public innovation policy.

The term innovation can be found in a number of different ways in the current literature. Joseph Schumpeter, the pioneer of innovation economics, considered innovation as part of a linear innovation process which contains the initial stage of invention, the stage of innovation and the final stage of diffusion. In this process, he regarded innovation as an invention that has been commercialized on the market by an entrepreneur! To understand this definition properly it is necessary to distinguish between innovation in the broad sense and innovation in the narrow sense. The first refers to the innovation process while the second refers to a single step within the innovation process. Schumpeter distinguished between product and process innovation, the opening up of a new market or new supply source and the carrying out of a reorganization. Further he made a basic distinction between incremental and radical innovations (Neveling et al., 2002). A radical innovation implies that a totally new product or process is developed which can transform existing markets or industries, or even create a new area of business. Incremental innovations on the other hand are small and often stepwise improvements of already existing products and processes. In practice they are much more common than radical innovations as no major organizational changes, investments or customer adjustments are required (Andersson et al., 2002). Although Schumpeters theories are already about 100 years old – he established them 1912 in his book called "Theorie der wirtschaftlichen Entwicklung" – most of the recent literature is still based on them. The subdivision of the process into invention, innovation and diffusion is basically still recognized, although this process is not explained as a linear one anymore, but as an iterative one. Furthermore some authors also include R&D and imitation in the process. Also Schumpeter’s distinction between radical and incremental innovation is acknowledged by modern literature as well as his detailed differentiation between the various sorts of innovation. However most authors nowadays concentrate exclusively on process and product innovations (Neveling et al., 2003, Brockhoff, 1999). For example the OECD describes innovation as "implemented technologically new products and processes and significant technological improvements in products and processes. An innovation has been implemented if it has been introduced on the market (product innovation) or used within a production process (process innovation)" (OECD, 1997, p.31). In this connection it is important to notice that we already talk about an innovation if the product or process is new to the enterprise, although it may not be new to the world (OECD, 1997). Similar definitions can also be found in a great variety of other books (e.g. Weber et al., 1999; Seidl, 1993; Meyer-Krahmer, 1989). Todaro finally describes innovation in his book about economic development as "the application of inventions of new production processes and methods to production activities as well as the introduction of new products" (Todaro, 2000, p.749).

Following the recent trend in literature, I will only consider the influence of public incentives on technological product- and process innovations, leaving aside all other types of innovation. Furthermore, also R&D and imitations will be included in the innovation process. Besides, it should be investigated whether different public instruments are used in support of incremental and radical innovations. A distinction between these two makes sense for two reasons. First, there are differences in the way they can be measured. Second, there are differences in the way the possible outcomes can be anticipated by the innovating company as well as by the government. While for incremental innovations it is quite easy to identify risks and establish the probability of success (not only for the company but also for the society), this is not true for radical innovations (Latour, 1999). Finally the following graphic should illustrate how the innovation process is described in modern literature. In the same way the concept will be used in this paper.

illustration not visible in this excerpt

Graphic 1: The innovation process today – innovation in the broader sense

Source: based on Brockhoff, 1999, p.36-37

As we can see the whole process normally starts with an idea about a project. This idea can come from entrepreneurs but also from another source, such as the government. If the chances for a technical and economical success of the project are assessed positively, green light will be given to start R&D. The result is either a technical failure or an invention. If it leads to an invention we can distinguish between planned and unplanned inventions. In the later case the invention is the result of serendipity effects, which means that the invention is a by-product of an R&D project that was carried out for another purpose (Maynz, 2001). The phase of implementation either leads to an economical failure or to an actual innovation. Depending on the market success of the product or process a slower or faster diffusion of the innovation takes place. Finally, this diffusion attracts the attention of competitors, some of which may decide to imitate the product or process. But this is not the whole story. As it was mentioned before, the innovation process is an iterative one. This means that you normally cannot simply go straight from the idea to the diffusion of the innovation. It is rather common to see the process as one where innovators have to go through many loops of trial and error. These loops can start at almost every phase of the process and they can lead back to almost every phase. Some of them have technical reasons, some are caused by consumer demands, others by activities of competitors. It would take much too long to discuss all kinds of loops possible during the process, but just to get an idea think about the recent trend in the market for mobile phones. Already for a long time the "innovation mobile phone" has been in the last phase of the innovation process. Nevertheless companies are regularly forced to jump back to the R&D phase in order to achieve (incremental) innovations. Examples are the recent implementations of WAP mobile phones and SMMS. But also loops before the phase of innovation can be possible, when for example internal tests show that the invention is still susceptible to defects or when consumers cannot be reached by the way the invention was commercialized.

To summarize there a two details that are remarkable in graphic 1. First it warns us not to forget about previous stages of the innovation process as we may be forced to come back to them. And second it shows that a public innovation policy cannot work efficiently without a supporting research respectively science policy, as R&D is a precondition for innovation. In the next chapter this fact will be given some more time and attention.

2.2 Innovation, technology and science policy

When looking through the relevant literature it is amazing how many different definitions of the various public policies can be found. This might be explained by the fact that innovation, technology or research policy must not be seen as isolated policies. They rather overlap and complement one another. Lundvall for example defines innovation policy as a mixture of those elements of science, technology and industrial policy that explicitly aim at promoting the development, spread and efficient use of new products, services and processes in markets or inside private and public organizations (Lundvall et al., 1997). Shyu is convinced that "science, technology, and innovation policy (in the narrow sense) are specific parts of what could be labeled more broadly as "innovative policy" (Shyu et al., 2001, p.230). As we can see from these examples it is common to see innovation policy as some kind of higher policy consisting of a core of own elements surrounded by components of the other policies. I will take up this idea in my paper. For a better understanding of how the various policies interact with innovation policy, the various supporting policies will be defined first. Then, it will be explained why innovation policy is more then just the sum of some components of these policies.

Science policy is defined as government-initiated policy which is concerned with the development of science and the training of scientists (Dodgson et al., 1996). Furthermore the OECD explicitly includes R&D activities to the area of responsibility of science policy (OECD, 2002). Technology policy consists of those measures which should guarantee respectively improve the standard of living in a society by supporting, developing and using new technologies (Bundeswirtschaftskammer, 1993). Industrial policy finally can be regarded as a policy which pursues a multitude of aims including increased output, short-term employment, better distribution and enhanced technological capacity (Bora, 1999). As we can see in graphic 2, science policy as well as technology and industrial policy supply important elements to the innovation process. Science policy has an influence on inventions as it is responsible for R&D activities. If we take a closer look at the definition of technology policy we can presume that it plays an important role in the phase of innovation and diffusion. Industrial policy supports the process in a more general way by creating a positive climate for innovations. Innovation policy in the narrow sense, finally takes into account the complexities of the innovation process and focuses more on interactions within the system. Dodgson describes the characteristics of innovation policies as follows. "They have the aim of improving the capacity to innovate of firms, networks, industries and entire economies. Innovation is a process which involves flows of technology and information between multiple actors, including firms of all sizes and public and private research institutes. Innovation policy's principal aim is to facilitate this interaction" (Dodgson et al., 1996, p.4). With this definition he already indicates the concept of National Innovation Systems, that will be explained in greater detail in the next chapter. Although I think that Dodgson definition is quite good and extensive I would like to add one of Sunil Mani. In one of his articles he wrote that "innovation policy is defined as a set of instruments and institutions which aid in the local generation of technology. This may also include adaptations of imported technologies to local conditions" (Mani, 2001, p.9). The reason for explicitly stating this description is the fact that it perfectly fits to the conditions of a developing country like Brazil, where the import of technologies plays an important role in the technological development.

Science policy Technology policy and Innovation policy

illustration not visible in this excerpt

Graphic 2: The influence of the different policies on the innovation process

Source: worked out by the author

To conclude this chapter it should be accentuated that the fact that innovation policy has to be seen as some kind of mixture of policies can lead to serious difficulties in implementing an appropriate innovation policy as the responsibilities for the various policies are normally divided between several different government departments (Rothwell et al., 1981). However, this problem will be discussed later in the paper.

2.3 National Innovation Systems

When analyzing the various public incentives that might be used in support of R&D and innovation it is important to notice that the government is not the only institution that has an influence on the innovation process. It is rather one of many, and the objectives of the other parties may either support measures taken by the government or weaken them. The first who discussed this fact in greater detail were Chris Freeman, Bengt-Åke Lundvall and Richard Nelson. They realized that R&D and innovation in an economy is influenced by a multitude of different institutions, which act and above all interact in a system they named the National Innovation System (or NIS) of a country (Freeman, 1997; Lundvall, 1992; Nelson, 1993). In this system the government with its policies is only one player of many. Besides, of course the companies and customers play an essential role, but also universities and other institutions of higher education, private research labs, providers of consultancy and technical services, and technology transfer institutions might be important. Finally, also the providers of financial and marketing services can exert an influence on the innovation process (John de la Mothe, 2001; OECD, 1999; Lundvall 1992). However, it should be noticed that because of historical and social reasons the importance of the various actors varies from country to country (or even from region to region), and in some countries or regions some institutions may not have any influence at all (Lundvall, 1992). As one already might have noticed, although national characteristics always play a role in shaping an innovation system, they do not only exist on a national level, but there can also be supranational, regional or local systems of innovation. The concept of Regional Innovation System (RIS) will be the important one for this paper, as the RIS of Espirito Santo will be used as one source of explanation why certain incentives or political instruments can be found in Espirito Santo.

3 Should the government support R&D and innovations?

The opinions on this question are divided. Some economists think that the public support of R&D and innovation is indispensable for the economy to work, while others reject any kind of public intervention in the market process from the start. Both opinions should be discussed in greater detail.

3.1 The arguments for public support

The common argument of those preferring governmental intervention in the process of innovation is market failure. In general terms, a market failure occurs when the free market left alone does not produce the welfare optimal outcome (Diederen et al., 1999). Necessary conditions for markets to yield Pareto-efficient outcomes include the absence of externalities and public goods, perfect competition, a complete set of functioning factor and product markets, and perfect information. These conditions are not completely satisfied in any economy, but especially in developing countries these market failures can be observed in a much broader scale (Ahrens, 1999). The most relevant market failure which justifies public intervention in the field of R&D and innovation is without any doubt the one of externalities. Nevertheless also problems concerning public goods as well as the degree of competition and the information available in a certain market are of relevance and should thus be discussed.

Although market failure is used as a standard justification for public support of R&D and innovation, it should not be the sole motivation for the government to intervene in the innovation process. It must also address system failures, as measures undertaken to prevent market failure cannot work efficiently with system failures still present (OECD, 1999). "System failures block the functioning of innovation systems, hinder the flow of knowledge and technology and, consequently, reduce the overall efficiency of R&D efforts". They can be the result of mismatches between the different components of an innovation system, such as conflicting interests of and incentives for market and non-market institutions (e.g. enterprises and public research institutions). But also inappropriate communication systems, as well as a lack of networking can be responsible for such failures (OECD, 1999, p.10). While each type of market failure described below calls for a specific way of public intervention, systemic failures require a broader range of initiatives. Therefore the government has to secure framework conditions that are conducive to innovation, such as a stable macroeconomic environment, an appropriate infrastructure, an appropriate education policy, or a public policy that concentrates on small and medium sized enterprises. Although these conditions may not directly influence the promotion of R&D and innovation, they are likely to have important indirect effects (OECD, 1999; European Commission, 2002). This shows once again, that we cannot only focus on what the government should do in respect to its R&D and innovation policy, leaving aside the influence of the other policies and private actors.

3.1.1 Externalities

Externalities in the field of R&D and innovation are also known as spillover effects. They can either take the form of price spillovers or knowledge spillovers. "Price spillovers occur when the increased value of new or improved products or services is not fully reflected in the price differential between the old and the new versions of the product or service." Therefore the innovator of the new product or service is not compensated for the portion of the benefits captured, but not paid for, by the user (Tassey, 1997, p.86). Knowledge spillovers resulting from the public good character of technological knowledge, which means that it is both non-exclusive and non-rival in character. The first implies that a user of technological knowledge cannot exclude others from the simultaneous use of this knowledge. The second means that the use of technological know-how by one company or person does not limit the use of this know-how by others. This leads to the following problem. As the production of technological knowledge costs money, a potential innovator will only be willing to invest in R&D if he thinks that he will be able to appropriate the economic and commercial benefits of the expenditures he incurred. Because of the public good character of technological knowledge, however, competitors of the potential innovator will be able to imitate and use the new knowledge without paying for it. Thus the potential innovator has to bear the costs of the R&D activities, but cannot, at least not completely, profit from the results. The outcome of the two spillover effects is a market that might not offer enough incentives to behave innovative and invest in R&D projects. That is why the level of innovations would be too low from a public point of view if the private market forces alone were to solve the problems discussed above (Hanusch et al., 1993; Marklund, 1994; Diederen et al., 1999). This is especially true for basic research as the general nature of its results attracts a broader group of potential adapters. Therefore the chances for spillover effects to occur are much higher with basic research then with applied research, as the results of the latter are only of value for a few other market participants (Overbeke, 2001). Nevertheless, in both cases public incentives are needed to overcome this market failure. However, it has to be noticed that externalities can justify a government intervention only in those cases where a project would not have been carried out without public support. There are also cases where market failures diminish the profits of the investing firm, but where still perfectly satisfactory returns can be obtained. In practice, however, it is often not easy to assess if a project will be carried out without public support or not. More attention will be paid to this problem in a later chapter. Besides, it still has to be shown that spillover effects can have a positive effect on the innovation behavior as well (Marklund, 1994; Diederen et al., 1999; Harhoff et al., 2000). Finally, it has to be noticed that markets do not automatically underinvest in R&D if left alone. On the contrary, in some industries an overinvestment in R&D might take place. This could be the case for industries where, for any reason, it is important to be the first on the market. Strong patent systems can play a negative role in this connection, at least if they imply too strong advantages for the patenting company. In this situation patent races will be more likely where each firm runs its own research program in the hope of being first to create and patent a new good or process. Besides, in order to be first the competing firms might accelerate their research programs at the cost of incurring additional expenses. All this results in an overinvestment in R&D which, from the society’s point of view, is as undesirable as an underinvestment since the extra money could have been spent on additional R&D and innovation projects (Jones, 1999).

A special form of externalities are network externalities, which "arise when the advantage of using a certain technology grows as the number of users of this technology increases" (Becher et al., 1995, p.315). In the case of network externalities it would make sense for the introducing company to keep the price for the product low in the early diffusion phases so that the number of users grows rapidly. The company then should get compensated for this initial low price by higher prices in later phases. A market failure may occur in this connection if the company expects the later market entry of competitors who just profit from the market development expenditures of the introducing company without having made their own contribution. In this case the introducing company will start with a price that is socially too high, causing a too slow diffusion of the innovation (ebenda).

3.1.2 Public goods

Another market failure can arise in connection with public goods. In this connection, the military forces, the health service, the education system or the protection of the environment are examples that are worth mentioning. Moreover, basic research belongs to the group of public goods as well. Incentives given by the private market would lead to an insufficient supply of innovations in these areas, as efforts undertaken by the supplier generally will not be rewarded adequately. Therefore public intervention is needed to prevent an undersupply of these socially important goods (Dreher, 1997; Diederen et al., 1999).

3.1.3 Market structure

In the discussion about the possible market failures occurring in connection with the market structure, it is important to notice that economists are still engaging in lively debates about the influences of market structure on R&D and innovation. Those economists following the Schumpeterian Hypothesis are of the opinion that industries with greater degrees of monopoly power should generate more innovations. First of all larger firms are in a better position to finance R&D projects and therefore innovations. They may be able to support a larger portfolio of R&D efforts, increasing the likelihood that it will develop an improved product or process. They may also have scale advantages in the R&D process (especially a large diversified business) as well as in the later production of the innovation. Finally, they may market innovations more effectively. The opponents of this theory, however, argue that a monopolist can have less incentive to innovate because in the absence of actual or potential competition there is little to be gained from engaging in a risky venture like R&D. Besides, large firms’ organizational structures may deter innovative activities (Nelson, 2002). In this connection empirical evidence suggests that there is no one answer to whether small or large firms are more innovative. The relationship between the market structure and the rate of innovation is rather sensitive to the type of industry being considered, the underlying knowledge base of the technology, and the specific phase in the industry life-cycle (Mazzucato, 2000). I will give some more time and attention to the analysis of these factors later in the paper. For now it should be emphasized that since a general relationship between the degree of competition and innovations cannot be proved empirically, it should not be the job of innovation policy to broadly prevent monopolistic market power although it represents a market failure. The government should rather decide from case to case if and how it makes sense to exert an influence on the market structure in order to support R&D and innovation.

3.1.4 Imperfect information

Potential market failure caused by imperfect information can occur at various stages of the innovation process. However, risk and uncertainty are of importance especially for the early phases of the process. Although the two terms are often used interchangeably in everyday speech, they have different meanings in economic discussions. Frank Knight was one of the first who realized that a distinction between these two terms is of importance for economic analyses. In his book “Risk, Uncertainty and Profit” he wrote that risk is present when future events occur with a measurable probability. Uncertainty, on the other hand, is present when the likelihood of future events is indefinite or incalculable. That means that risk is measurable while uncertainty is not (Knight, 1921). Although both are the result of imperfect information they influence R&D decisions in different ways. If the company is able to ascertain the probabilities of a positive or negative outcome, that is the risk of an R&D project, it has an objective basis to decide for or against an R&D project. With uncertainty a company does not have such a basis. Instead of basing its decision on facts, the firm respectively its managers will base their decision for or against an R&D project on past experience and expectations for the future (Keynes, 1921). The problem in this connection is that personal interests of the managers and lobbying might influence the decision process which might have a negative influence on the R&D activities of a company (Gollier, 2003). The risk aversion of managers clearly influences the innovative performance of a firm in a negative way as every investment in R&D is at least a risky undertaking. On the other hand, managers care about the growth of their firm as a manager’s income, personal power and prestige are directly related to the firm size. This influences the innovative activities of a company in a positive way as more R&D increases the chance of developing new products which in turn leads to a larger sales volume and a faster growth. In the end, it is difficult to say if the subjective decisions of managers lead to more or less R&D activities inside a firm compared to when the decision is based exclusively on an objective risk profile. The result depends on the relative risk aversion of the manager and on how important the fast growth of the firm is to him (Czarnitzki, 2001).

Normally, it is possible to calculate the risk for some steps of the innovation process. However, most of the time the success of the whole R&D project is uncertain. For instance, it is not sufficient to know the probabilities of a technical success of an R&D project. For the success of an innovation project it is also of importance what the competitors are doing. If a firm is not alone in doing research in a particular field, several occurrences can endanger the success of the innovation project and therefore make it uncertain. First, there is the possibility that a competitor is rewarded a patent for its research efforts because it was the first to obtain a patentable result. This means that the research already done by the firm becomes useless as it is not allowed to use its knowledge. Second, even if the company is able to get a technology commercialized in time there is the risk that a competitor enters the market with a superior technology before the firm was able to make up for the expenditure on R&D. An ever greater number of technology-based competitors makes such scenarios more and more possible. This, together with an ever increasing capital intensity of R&D makes it for some firms harder and harder to align their R&D decisions with their risk-diversion policy. This is especially true for small and medium sized enterprises. Because of their relative low financial reserves, a failed innovation project might be fatal to them. Therefore, the risk preferences may shift to lower tolerance levels, and decisions are made to reduce total R&D or to shift investment to short-term R&D projects with a lower risk at the expense of long-term projects with potentially higher social rates of returns (Tassey, 1997; Branscomb, 2001, Straathof, 2002).

But risk and uncertainty cannot only be found inside the company. Also financial institutions and capital markets may lack the information and technological knowledge to properly assess the uncertainty or risk attached to R&D and innovation projects and therefore may be overcautious in providing the required finances. Therefore, too much of the financial risk will be left to the potential innovator, which again might deter companies from carrying out R&D projects or other innovative activities (Diederen et al., 1999).

Another market failure caused by imperfect information may occur in the phase of diffusion. If the existence and the current characteristics of an innovation are not general knowledge, it can happen that other firms undertake R&D for something that already exists – that is they try to reinvent the wheel. These unnecessary replications of R&D and innovations, however, represent a loss to social welfare, as the financial resources could have been used to invent products and processes not existing yet (Stonemann, 2002). Some might object now that it was argued in the chapter about externalities that the use of technological knowledge of an innovating firm by others might reduce the willingness of the innovator to undertake further R&D projects. In this connection it is necessary to distinguish between two aspects. The question if the knowledge developed by an innovator is used by others and the question if the innovator is paid for this or not. Only the later poses a problem to the innovation process, and it still has to be discussed how the government can intervene to solve it.

3.2 The arguments against public support

Although also neoclassical economists admit that market processes are imperfect, they argue that markets still provide adequate mechanisms for efficient economic coordination. Rather governmental intervention in the market is regarded as inefficient. Consequently, occurring imperfections are frequently interpreted as an outcome of wrong public policies, that means as government failure (Rothwell et al., 1981). The reasons for government failure are of two different kinds. First, the government can never be as in tune with market forces as industry participants, which means that the entrepreneurs are always in a better position to make rational choices about alternative investments in R&D and innovation. This is especially true for fast-changing markets. Second, government policies are normally influenced by several political forces (like political rivalries or election cycles), by lobbying and sometimes even by corruption. In these situation politicians do not decide on an economic basis anymore and thus distort the market mechanism (Dai, 1996; Dreher, 1997; Ahrens, 1999).

Therefore the opponents of governmental intervention think that the role of the government should be limited to interventions in case of substantial market failures, for instance when markets do not exist or when strong externalities are prevalent. Otherwise it should only prevent system failures by providing a stable macroeconomic environment and a market-oriented legal infrastructure (Ahrens, 1999). Some arguments of the neoclassical theory should be considered when choosing the appropriate instruments to support R&D and innovation, but the theory as a whole is a rather bad approach to achieve satisfying outcomes in the field of technological innovation, especially in a developing country like Brazil where market failures can be observed on a much broader scale than in any developed country. Besides, it is also questionable if the Brazilian market (especially the capital market) is strong enough and provides adequate mechanisms for self-regulation. That is the reason why in a developing country like Brazil extensive public support of R&D and innovation is of need. However, it is also true that the use of the innovation policy instruments is not always an unproblematic one. Therefore, as it has already been said, the government has to weigh up the advantages and disadvantages of any incentive it intends to use before doing so. This means it should try to find out to what extent a market failure makes an intervention necessary and whether or not a certain policy measure truly leads to an improvement. In addition, it has to compare the costs of an intervention with the estimated resulting social welfare gains. Finally, it also has to pay attention to possible side effects (Becher et al., 1995). If the government does all this, the losses from a government failure should, however, be much smaller than the losses from a market failure. In order to allow such a judgment, I will use the next chapter to explain the various instruments existing and to analyze their strengths and weaknesses.

4 The innovation policy tools

When dealing with the recent literature it is remarkable upon how many different instruments the government can fall back when it intends to influence the innovation process. In order not to lose track of things the various instruments existing will be grouped together. The most important distinction can be made between financial and non-financial measures. To the first group we count public funding, venture capital, and tax incentives. Consequently, all the other instruments available are belonging to the group of non-financial measures. Among others, those are the support of technology transfer, public procurement, public laws and regulations, industrial standards, the promotion of an innovative infrastructure, public policies on human resources development, or the policies promoting the co-operation between the various actors of a NIS. Table 1 at the end of the chapter will give a complete overview over the instruments discussed in this paper (Meyer-Krahmer, 1989; Diederen, 1999; Mani, 2002).

Furthermore, the government can either try to influence the supply side or the demand side of the innovation process. On the supply side it attempts to push new products or services on the market by supporting the innovative behavior of firms. This can be done for example with financial measures like subsidies or with non-financial measures like public information services. On the demand side the government aims to pull innovations through the innovation process by creating certain market needs. For example, it can do so with the help of public procurement or industrial standards. In practice, it will depend on the phase of the industry life-cycle as well as on other factors if more public support is required at the supply or at the demand side of the innovation process. However, the government should always take care of both (Rothwell, 1981; Dodgson, 1996; Meyer-Stamer, 1996).

Finally, public instruments in support of R&D and innovation can be of an indirect, indirect-specific and direct nature. The aim of indirect measures is to create a supporting climate for encouraging general innovative activities, without discriminating between specific technological areas. In contrast to this, indirect-specific measures support projects of either a certain technological area or area of use. Direct measures, finally, concentrate on particular projects with specific targets (Dodgson, 1996; Dreher, 1997). In this connection it should be noticed that almost every instrument can be utilized either in an indirect, indirect-specific or direct way depending on how it is employed by the government. However, it makes a great difference if a certain instrument is used directly or indirectly. The more direct the nature of a certain instrument employed is, the more influence the government will have on the decisions of the company, and therefore the greater the risk of government failure will be. Besides, the more direct the nature of the support, the less widespread the benefits will be. On the other hand, it is said that a more direct public support helps to prevent free-rider effects, as the government can collect more information about the companies and projects it intends to support. Free-rider effects occur where companies receive public support for R&D projects only because they fulfill certain criterions, also if the projects in question would have been carried out without public support as well (Meyer-Krahmer, 1989).

The grouping of the various instruments is illustrated in table 1. In this connection it should be differentiated between fiscal and non-fiscal instruments respectively between those instruments rather working at the supply side and those rather working at the demand side. In the latter case there are also some instruments working equally well to support the supply side as well as the demand side of the innovation process. They will be assigned to both groups.

illustration not visible in this excerpt

Table 1: The instruments of innovation policy

Source : worked out by the author

After getting a general idea of the various instruments existing, the following chapters should be used to discuss each of them in greater detail. It will be gone into the question of if and how a certain measure can be used to reduce market and/or system failures. At the same time it will be discussed which problems or risks might be connected to them. In this connection it will be devoted some more time to financial measures as they have attracted much attention and analysis in the recent literature. However, it still has to be surveyed if this interest in the financial instruments of the theorists is also reflected in the innovation policies of developing countries, as especially for them also non-financial measures are becoming more and more important.

4.1 Financial measures

4.1.1 Public funding

Although public funding for technological innovation is an instrument that can be employed at all stages of the innovation process, it is typically used to support R&D. The most common form of public funding are grants, which means that the government agrees to bear an agreed proportion of the costs of the innovation project (Aichholzer et al., 1994). Besides, also soft loans can be counted to this instrument. Those are loans with an interest rate below the market rate or with other favorable conditions. Finally, also loan guarantees for innovation projects represent a possible form of public funding. In this connection the government acts as a guarantor to the bank in case that the project fails (Diederen et al., 1999).

The strengths of the instrument

Public funding can have an effect on two different market failures. First, grants can be used to reduce the negative effects of externalities. As it was said before a company will only be willing to invest in R&D if it thinks that it will be able to appropriate the economic and commercial benefits of the expenditures it incurred. Grants do not have any influence on the fact that competitors of this company will be able to imitate and use the new knowledge without paying for it, but they reduce the expenditures made by the company and therefore close the gap between the costs and benefits of a project. Therefore, receiving a grant may turn an unprofitable project into a profitable one to be pursued by a company. Considering the way grants are working, it is not hard to understand that they play a vital role in supporting basic research as well as in promoting specific technological areas or projects of higher political interest like environmental conservation or national defense. In these cases the gap between the costs and the commercial benefits is very wide, and therefore the government has to subsidize the projects to a great extent, sometimes up to 100 percent (Meyer-Stamer, 1996).

The second market failure that can be reduced with the help of public funds is the one of risk. While grants and soft loans reduce the risk inside the company only, loan guarantees also have an influence on the capital market. Inside the company public funds may help to bring certain innovation projects in line with the risk-diversion policy of a company. This is especially true in the case of technologically ambitious projects, but also when we are talking about small and medium sized enterprises. Without public funding, normally, the latter are not able to carry out more than one or a few innovation projects at the same time. However, the impact of a failure of a project is much worse if it is the only one than if there are a lot of others that still can win the race. In this sense, public funding reduces the risks for small and medium enterprises by giving them the financial possibilities to carry out more projects at the same time. Furthermore, as it was described before, financial institutions and capital markets may be overcautious in providing the required finances as they lack the information and technological knowledge to properly assess the risk attached to an innovation project. With loan guarantees the government can eliminate the perceived risk of financial institutions by assuming the risk attached to the innovation project (Bundeswirtschaftskammer, 1993).

Finally, public funds cannot only be used to reduce negative externalities but also to increase economic benefits caused by positive externalities. First, the learning and know-how gained in subsidized projects might also spill over to other current or future projects of the company thereby enhancing their chances of success (Lach, 2000). Furthermore, it is expected that as the government makes more R&D money available to an industry, firms increase their own R&D spending, also in the hope of capturing more government funds (Mani, 2002).

The weaknesses of the instrument

One big weakness of public funding is the fact that it represents an extremely cost intensive instrument, especially when it is used to promote basic research. Therefore it is normally used as either a direct or indirect-specific measure, that is to support only a few technological areas or projects, simply because the government lacks the financial resources to give this kind of support to the masses. In this connection, the government typically concentrates on so called key technologies like new materials, biotechnology or information technology. However, it is questionable if the government is good at picking the right technological areas or projects. This is because it might not have sufficient insight into the market dynamics and therefore overestimates the potential of a chosen technological area or project. In the end, it is possible that the government supports "second best" areas or projects at considerable opportunity costs (Bundeswirtschaftskammer, 1993; Rothwell et al., 1981). To reduce this risk, the government should reserve at least a small part of its funds for initiatives from innovators working in non-priority areas, as their work might lead to new ideas and even new ranges of technologies which may offer opportunities not foreseen by the government (Aichholzer et al., 1994).

Another problem the government has to face when using public funds is the fact that it is difficult to estimate the actual gap between the expenditures necessary to carry out an innovation project and commercial benefits resulting from it. That means that it is difficult to lay down how big the financial support has to be in order to give a potential innovator enough incentives to carry out a project. First of all, it cannot be ascertained how big the actual expenditures and commercial benefits will be. However, this considered by itself should not pose a big problem as it is only important on how the company assesses these numbers. What causes the problem in this connection is the fact that companies usually are not honestly reporting these numbers, but that they are trying to state their expenditures as big as possible while describing the expected benefits as small as possible in order to receive the maximum amount of support possible. This leads to a situation where public funding partly or even fully crowds out privately financed R&D, which means that public funding substitutes for company funding instead of augmenting or complementing it. Besides cheating of the companies, there is the problem that the government might be overcautious when selecting projects to subsidize in order not to waste public funds and, therefore, tend to support projects with higher success probabilities, that is projects that are likely to have high private rates of return. Such projects, however, would have also been financed by the firm suggesting that the R&D subsidies are indeed superfluous and may be crowding out private R&D resources. In both cases, the problem can be reduced if public funds are used as a direct measure. Like this the government is in a better position to acquire enough information about a project and a company it intends to support in order to be able to assess if the project would not be carried out in the absence of public funding. If public funding is really required, then the acquired information will help the government also to audit if the presented figures can be correct. However, cheating is still possible (Lach, 2000; Meyer-Krahmer, 1989). Therefore, the government should use a special mechanism where the initial funding level is conservative while future increases depending on how the R&D project develops are possible. This takes into account that the more advanced a project is the better its expenditures and benefits can be assessed (Tassey, 1997).

Finally, it should be avoided to fund certain companies too often as this can have a bad influence on how these companies perceive the role of the government. Therefore, it can lead to a subsidization mentality of entrepreneurs which shows in an inactivity and a wait-and-see attitude. Furthermore, it can lead to a situation where companies seek profits increasingly through political lobbying and government subsidies and not through inventing, producing and selling (Rothwell et al., 1981; Becher et al., 1995).

Final considerations

The following figures 1 and 2 should illustrate how to the government, at least theoretically, can determine the rate of subsidy necessary to maximize social welfare. Besides, it is also possible to tell to which extent the subsidy expanded R&D efforts and to which extent the public subsidy displaced private expenditures on R&D.

illustration not visible in this excerpt

Figure 1 and 2: The effects of an R&D subsidy

Source: based on Dowrick, 1995, p.47

In the two figures the line ABC respectively ANBNCN represents a company's marginal returns to R&D while ADE represents social marginal returns. Line KL represents marginal costs of outlay on R&D. The optimal outlay on R&D for the company is given by the intersection of the line ABC respectively ANBNCN with the KL-line. In this point the marginal costs of an additional outlay on R&D equals the marginal returns of an additional outlay on R&D. This leads to an outlay of X1 respectively X1N, which in both cases lies under the socially optimal level. To achieve this socially optimal level the government has to subsidize the R&D activities of the company. As it can be seen from the two figures, the company's marginal returns to R&D play an important role in this connection. In the first figure the company's marginal returns to R&D line is less elastic then the one in figure 2. Therefore, to achieve the same optimal outlay on R&D (X2), the public subsidy per dollar of expenditure has to be much more extensive in the first case than in the second one (compare P in figure 1 to PN in figure 2). As a consequence also the crowding out of privately financed R&D by public funding is much bigger in the first case (compare the rectangle MCBK to the rectangle MNCNBNK). Finally, the more elastic the company's marginal returns to R&D line is, the bigger will be the increase of total private outlay on R&D. Therefore, the proportion P:O of the first figure is smaller than the proportion PN:ON of the second figure (Dowrick, 1995). What should be learned from the two figures above is that under the assumption of known company's or industry's marginal returns on R&D, the government should rather fund those companies or industries with more elastic marginal returns to R&D lines, as they respond more favorably to public subsidies. However, it is questionable how easily the degree of elasticity can be ascertained in practice.

4.1.2 Venture capital

Venture capital "is an equity form of investment in a technology-based firm at its early stage of development. In addition to providing the much-needed risk capital, the venture capitalist also renders a fair amount of value-added support to the investee" (Mani, 2002). The reason for the latter can be found in the fact that the venture capitalists in exchange for all his services demands a preferred equity share of the new venture, along with favorable upside and downside investment protections. Therefore, it is in his own interest that the venture is developing well (Callahan et al., 2003). There are two ways this instrument can be employed by the government. Either it provides venture capital itself, usually by participating in a company through government-owned investment companies, or it tries to stimulate the supply of venture capital from non-public sources. It can do so by providing tax incentives to those investing in small firms or venture capital funds, or it allows institutions such as insurance companies or pension funds to invest in venture capital (Mani, 2002; Diederen et al., 1999).

The strengths of the instrument

First, venture capital can help to overcome market failures caused by imperfect information in two ways. First, a project financed by venture capital can reduce the personal risk, as the inventor can be sure that no loss would be personally incurred to him. Second, imperfections in the capital market are reduced. It is likely for innovators to have sufficient understanding of the market they are operating in to assess the risk of an innovation project they intend to carry out. However, financial institutions usually do not posses this ability. Therefore, it may well be that the capital market reacts overcautious and that it becomes difficult for entrepreneurs to rise high levels of funds through traditional debt financing. This is especially true for firms that want to carry out riskier projects where a future success is not obvious, as well as for small and medium sized companies, as those are operating in small or niche markets which are less understood by the financial institutions (Callahan et al., 2003). Besides, this problem is also relevant for early-stage companies, as the innovative idea behind a product is usually untested, the period of time for a return on investment to be realized is comparatively long and the capital infusion required may be big (Pennsylvania Economy League, 1998). As mentioned before, public loan guarantees are a way to solve these problems. However, also venture capital can do this job, as it provides high levels of funding to opportunities with high risk or uncertainty (Callahan et al., 2003).

Another strength of this instrument is the fact that it reduces system failures. For the reason stated above many venture capitalists do not only offer risk capital to the companies, but they normally also render value-added support. They might provide specialized knowledge of a particular industry or facilitate the access to a network of contacts that may include seasoned managers, partners, and customers. Furthermore, they might also provide control and expertise to the company. In the end, this well-managed and well-financed company will offer an ideal breeding ground for R&D and innovations (Callahan et al., 2003).

The weaknesses of the instrument

Public venture capital can lead to similar problems as public funding. First of all, quite a few financial resources are needed in order to use this instrument, which considered by itself might already pose a problem to some governments, especially when we are talking about developing countries. Second, this implies once more that the government will have to use venture capital as either a direct or indirect-specific measure. Again, in this connection it has to be questioned if the government is good at picking winners.

Some may argue now that in the case of venture capital, the government has the possibility to delegate the job to market participants, to venture capitalists, who should have enough insight into the market to decide which projects are the very promising ones. Actually, they can do this. Nevertheless, the government would be unwise to exclusively concentrate on stimulating the supply of venture capital from non-public sources. This is because venture capitalists do have enough experience and knowledge to pick the winners, but they rather care about their own than about social welfare. Therefore, they strongly prefer particular industries with great potential for fast and substantial growth, like software. This can have two negative consequences. Industries with high social rates of return but which do not offer high growth rates might not be promoted (Callahan et al., 2003). Besides, as Tredennick puts it, "if you ride the crest of a fad, you've got a good chance of getting funded. If you have an idea that is too new and different, you will struggle for funding" (Tredennick, 2001, p.2). This means that maybe no attention is paid to some socially valuable ideas or projects which would result in welfare increasing innovations. To overcome this problem, the government should use a diversified policy. First, it should stimulate the supply of venture capital from non-public sources to provide venture capital for mainstream industries. Second, it should see venture capital as what it is, namely as risk capital, and use it to support new and different ideas or industries with yet unknown opportunities.

Final considerations

Apart from the discussion on how venture capital might influence R&D and innovation in the receiving company, it is worth analyzing how venture capital effects the innovation process of incumbents companies if they are not the receiver but the supplier of venture capital. In this connection it can be shown that corporate venture capital investments can result in knowledge spillovers the incumbent firm can use to overcome its inability to internally generate innovations (reasons for this inability may be found in stiff organizational structures or a lack of specific knowledge). "By investing in a new venture, incumbent firms may be able to learn about new technologies or practices." With this knowledge, they may be able to invent new products or diversify into new lines of business (Dushnitsky et al., 2002, p.6). In this connection it is important to notice that the incumbent firm does not steel the innovative idea from the venture it finances, but that it rather bases its innovations on what it learned during the support of the venture. (Dushnitsky et al., 2002). This means that as a result of venture capital new innovations are not only possible in the supported firm but also in the supporting firm. Therefore, it represents a process that the government should heavily promote.

4.1.3 Tax incentives

Tax incentives allow companies to reduce their tax bill as a reward for carrying out innovative activities. That means that the government may permit the company to deduct a part or even the entire expenditures connected to R&D projects from the taxable income. In some cases, an extra tax allowance is made, which means that an amount even greater than what is spent on R&D is allowed to be deducted from the tax base (Mani, 2002). Furthermore, the government may allow an accelerated or free depreciation made in machinery, equipment or building, used exclusively for innovation activities. Free depreciation means that the company decides how it wishes to depreciate the innovation capital expenditures, which normally means that it writes off immediately. Accelerated depreciation is equivalent to a shorter life of capital expenditure for tax purposes than of economic purposes. Both, however, only lead to a delay in tax payments and not to a tax reduction. Finally, there are also indirect incentives in the field of corporation tax. For example, the government can allow the company to carry forward a negative tax basis. Those interested in other indirect measures not discussed in this paper should read "Innovation papers No. 19" from the EU (European Commission, 2002b).

The strengths of the instrument

To begin with, tax incentives have some advantages over the other financial instruments. By far the most important one is that they entail less interference with market mechanisms, as they allow the markets to determine the allocation of R&D investments across sectors, firms and projects. Companies can decide on how they want to spend their R&D budgets and where investment is required most, rather than this being determined through a bureaucratic central authority. They are free to respond to real market demands as opposed to government-created demand. Therefore, tax incentives help to reduce the risk of government failure. At the same time they are able to overcome the consequences of negative externalities as they, like public funds, are reducing the costs inherent to innovation projects. Tax incentives also involve a lower administrative burden compared to other financial instruments. Therefore, the costs of tax incentives in terms of time and money are reduced for both the company and government alike. There is little if any prior application procedure the company has to undergo. To be sure that the incentive will be awarded it simply has to fulfill specific predetermined criteria. The government, for its part, solely needs to control if certain activities of the company meet the qualification criteria set out before. Finally, tax incentives are also more predictable and more stable as they are less subject to legislative change and yearly budget allocations. This also makes it easier for companies to correctly plan their future innovation activities (Mani, 2002; OECD, 1996 and 2002b; European Commission, 2002b).

Besides reducing the negative effects of externalities, tax incentives can also be used to lessen the consequences of risk. The government can do so by allowing companies to carry back or forward a negative tax basis. For companies diversifying into new technological areas or for high risk innovation projects, the opportunity to compensate for losses within past or future profits may provide sufficient assistance to cause a company either to enter "unknown territories" or to start a more risky innovation project that it otherwise would not carry out.

The weaknesses of the instrument

While the fact that tax incentives are available to every company fulfilling the predetermined criteria represents an advantage to entrepreneurs, it might pose a problem to the government. This is because it might be possible that a company classifies routine research expenses as innovative or that it relabels routine expenditures such as quality control and testing as R&D expenditures and then claims tax incentives. Furthermore, it makes it even more difficult to ascertain if public support expanded R&D and innovation activities or only displaced private expenditures on R&D (Mohnen, 1999; Mani, 2002). In this connection, there is also the problem that a government can hardly evaluate in advance how much the use of tax incentives as an instrument will cost. The government will have a declared budget, but as claims are received "only after it would already be too late to complain" there will always be a level of uncertainty regarding total actual cost. To overcome this problem the government might lay down limits of fiscal incentive, which means that R&D expenditures exceeding a certain level cannot be deducted anymore from the taxable income (European Commission, 2002b).

Furthermore, there is the problem that start-ups and SMEs often cannot be reached by tax incentives. They often do not have enough taxable income form where they could deduct all their R&D expenditures. Therefore, it would be important to allow them to carry forward a negative tax basis. In this way the government makes it possible for those companies too to deduct their entire R&D expenditures (European Commission, 2002b).

Final considerations

Public funding and fiscal incentives should not be seen as alternative instruments but rather as complementing ones. While the first one is a direct and target-oriented measure, the latter is an indirect measure that works through the market mechanism. That means that with the former instrument the government is able to assist key projects, industries and technologies, while with the latter it is able to promote a general increase in technological innovation expenditure (Tassey, 1997; European Commission, 2002b). Therefore, the government should not decide on whether to use one or the other instrument but rather on how many financial resources should be allocated to each of them. In the case of developing countries, the government is in a dilemma. On the one hand, it needs to support basic research which can be done best via funding. On the other hand, it knows that fiscal measures are necessary to encourage general innovation expenditure in companies, especially in countries with a poor innovation policy tradition as it is the case for many developing countries.

4.2 Non-financial measures

4.2.1 Public procurement

Public procurement and public technology procurement are typical examples of an innovation policy instrument working at the demand side. While public technology procurement can have an influence on the first stage of the innovation process, simple public procurement influences the stage of diffusion. In the first case, the government initiates the process by placing an order for a product or a service that does not yet exist. That means that the government indirectly instructs a certain company to carry out an R&D project to invent the demanded product or service. Normally, after its development, the product or service is continuously ordered by the government. This allows the company to cover their investment made in R&D (Dreher, 1997; Edquist, 2001). Besides, simple public procurement influences the diffusion process of a product or service. By deciding on a certain kind of technology, it can have an influence on which technologies are used by an industry (Aichholzer et al., 1994).

The strengths and weaknesses of the instrument

The fact that public technology procurement is usually connected to later guaranteed public orders helps to reduce the negative effects of both externalities and imperfect information. While this measure is not able to (and often does not intend to) reduce knowledge spillovers, it increases the expected profits and therefore closes the gap between the expenditures and commercial benefits of an R&D project. This is because the company has a guaranteed future profit by doing business with the government. Furthermore, it also reduces uncertainty and risk as the company knows more about future market conditions (Dreher, 1997). Besides, public procurement is perfectly suited to pursue social aims. For example, the government can, through its own demand, influence the choice between different kinds of energy sources and thereby which energy conversion technologies are used. Finally, the results of R&D and innovation activities stimulated by public procurement might spill over to other companies or industries. Thereby, public demand does not only support companies selected to carry out a project but the economy as a whole (Edquist, 1996; Aichholzer et al., 1994).

First of all, as public procurement is almost per definition used as a direct measure it can be questioned, as in the case of public funding for example, if the government is good at choosing the right technologies. Besides, it is questionable if all public decisions are taken on an objective basis or if lobbying and personal preferences of single public agents also play a role in the procurement process (Dreher,1997). Furthermore, the absence of market power on the side of the government may reduce public influence in certain areas. Finally, also a lack of consensus in the government, resulting in frequent changes in the procurement policy, can weaken the positive effects of the measure (Rothwell et al., 1981).

Final considerations

First of all, it has to be understood that guaranteed public orders for newly invented products or services are an important factor of public technological procurement, but nevertheless it is crucial that the procurement process involves a substantial element of competition among potential suppliers. This requirement can be met if the public procurement initially takes place through open bidding (Edquist,1996). Furthermore, it should be emphasized that the influence of public procurement is dependent on the relation between market and government power. Therefore, it makes a difference if the government is the only buyer or one of many. But also the market structure on the supply side can put constraints on the possibilities of a government procurement policy. Therefore, it makes a difference if there is only one or a few companies able to support the demanded product or service or if there are a lot. Who wants to know more about these relations should have a look at Rothwell (Rothwell et al., 1981).

[...]