CROATIAN INNOVATION POLICY AND ITS EFFECTS
Discussion Materials for the GDN Conference
Prague, August 14-17, 2004
Draft version
- please do not quotewithout permission -
Project coordinator: Domagoj Račić ()
Research team:
Zoran Aralica ()
Katarina Bačić ()
Domagoj Račić ()
Dubravko Radić ()
Consultant:
Slavo Radošević ()
PART ONE
RESEARCH PROJECT DOCUMENT
Croatian Innovation Policy and Its Effects
- Abstract-
Recently, the relationship between innovation and economic growth has become widely recognised. Within transition countries, that is facilitated by both exhaustion of growth and productivity improvements based on non-investment reallocations and the integration into the European Union, which states the development of a knowledge-based economy as a crucial policy goal. The research project demonstrates how a national innovation policy can facilitate restructuring, technological advancement and economic growth in Croatia, and assesses the achievements in that regard so far. The project positions innovation policy within the general economic policy framework and analyses the innovation-facilitating institutional, fiscal and financial mechanisms implemented by the policy makers. Moreover, summary of the level of innovative activities in Croatia, and the evaluation of the effects of the emerging innovation policy are provided. Finally, the project deals with the barriers to innovation, and further challenges and policy options that could lead to a more effective innovation policy. The project involves two papers with different but complementary methodologies. The first one assesses the treatment effects of selected policy instruments and other determinants of innovation activities by using a probit model that estimates their contribution to innovation propensity, controlling for the sample selection problem.The second one involves a descriptive analysis of innovation policy and its basic outcomes, complemented by relevant statistics and indices (some based on the European Innovation Scoreboard, some purposively constructed).
Croatian Innovation Policy and Its Effects
a) Research questions:
1. What are the key features (scale, patterns, and determinants) of innovative activities of Croatian manufacturing and service enterprises and what are their implications for policy?
2. To what extent and in which ways can an innovation policy facilitate innovative activities and contribute to restructuring, technological advancement and economic growth in Croatia and what have been the achievements in that regard so far?
b) The issues addressed by the research project:
• The relationship between innovation and economic growth
Recent years have seen the emphasis of the relationship between innovation and economic growth. At the macroeconomic level, innovation tends to contribute to the accumulation of capital, and growth of employment and multifactor productivity (OECD, 2001). However, the relationship between research and development (R&D) expenditures and growth is not straightforward. Although significant, aggregate R&D explains limited part of variation of growth. This suggests that the factors such as allocation of R&D resources, and the mechanisms of creation, dissemination and commercial exploitation of knowledge matter fundamentally. These factors are significantly influenced by the national innovative capacity (Stern, Porter and Furman, 2000), which covers the ability of a country to produce and commercialise a long-term flow of innovative technology. R&D or ability to generate new knowledge is only one component of broader concept of national innovative capacity. Among other issues, national innovative capacity depends on the strength of a strong common innovation infrastructure, i.e. cross-cutting factors that contribute to innovation throughout the economy. Hereby the crucial factors include the science and technology policy, mechanisms for the support of basic research, and the accumulated stock of technological knowledge including diffusion and utilisation of the existing knowledge. Consequently, government policies can play an influential role in the facilitation of innovation - notably through support for R&D, education and labour market policies, entrepreneurship support, and the promotion of interaction among different organisations within the national innovation system (including the research institutions, business firms and government alike).
• Innovation policy in transition economies
The realisation of the role of innovation policy has traditionally been stronger in more advanced economies. The transitional countries in Central and Eastern Europe (CEE) have initially focused their policy efforts on macroeconomic issues and measures (cf. Bučar and Stare, 2002). However, the realisation of the importance of innovative activities and innovation policy has recently been facilitated by two key factors. Firstly, there has been an exhaustion of growth and productivity improvements based on defensive restructuring and non-investment reallocation of resources (cf. Mickiewicz and Radošević, 2001). Since competition on the basis of low wages is an unfavourable and unsustainable strategic option for most of these economies, their long-term competitiveness requires technological advancement and the development of innovative capacities. Moreover, most of these countries are being integrated into the European Union. The EU not only states the development of a knowledge-based economy as a crucial policy goal for its current members, but also requires from the candidate countries to demonstrate the same orientation. Namely, innovation and increased technological change in new member states are viewed as fundamental to their economic convergence with current members and the cohesion of the enlarged EU. Consequently, maintenance of productivity and GDP growth require new mechanisms for supporting innovation and industrial upgrading (cf. EC, 2001: 11). Furthermore, the EU is already the main foreign trade partner and the source of FDI for transition economies during the pre-accession period, which reinforces the need for maintaining and improving internal and external competitiveness.
In terms of relative wages, Croatia is faring badly in comparison to the rest of CEE, which undermined the competitiveness of several traditional export-oriented sectors (e.g. textiles and apparel industries). On the other hand, retarded levels of technological capacity and product and process innovation have not provided an alternative route to competitiveness. The examples of internationally competitive innovations have been sporadic (e.g. the pharmaceutical company Pliva) and they have rarely induced strong spillover effects. Moreover, inadequate factor markets (i.e. inflexible labour market and underdeveloped capital market) and insufficiently supportive policy mechanisms have even encouraged dislocation of certain activities to other CEE countries. Deteriorating export competitiveness has been observed both in the EU and CEFTA markets.
• Croatian innovation policy within its economic policy
Therefore, the necessity for the development of innovative activities supported by adequate innovation policy is paramount for the catching up in terms of technological advancement, product and process developments and the resulting competitiveness of the Croatian economy. However, the realisation of this need by researchers and policy makers and the development of adequate policy mechanisms have been relatively slow and occasionally inconsistent. The economic policies mostly focused on macroeconomic stabilisation, reforms of the financial system, taxation and the pension system, and liberalisation of trade and exchange regimes. When it came to microeconomic issues, the policies have often been related to privatisation and restructuring of existing enterprises; even in these areas the success has been mixed due to political influences and weak institutional frameworks. Consequently, not enough emphasis was given to the issues of new enterprise development, promotion of innovative activities, creation and effective functioning of interfaces between research community and industry, or the facilitation of integration of innovative enterprises into local, national and global industrial networks. Moreover, due to inadequate investment promotion policy, the FDI inflows have predominantly occurred through privatisation of existing firms for market seeking reasons, mainly in three sectors (pharmaceutical industry, banking and retail), and they have not resulted in significant technology transfers or spillovers. More recently, there have been improvements within the area of the issue of enterprise development: new credit lines have been secured, and better technical assistance to entrepreneurs provided, which had noticeable effects on the performance of the SME sector and led to wider acknowledgement of its role in the economic growth and job creation.
• Institutional and regulatory frameworks: structures, relationships and challenges
The project analyses the institutional framework for innovation. The focal institution for science and technology (S&T) and innovation policy issues has been the Ministry of Science and Technology, which has been involved in organisation and financing in the areas of scientific research, technological development, higher education, development of activities related to information technology, as well as international scientific and technological cooperation. The role of other relevant governmental institutions (relevant ministries, implementing agencies etc.) and their stakeholders (entrepreneurship support centres, business incubators, technology parks, professional and business associations) will also be assessed and evaluated. The project also analyses crucial legal, fiscal and financial provisions and instruments that have been established to promote innovation in enterprises, as well as the key business regulatory issues, such as the factors hindering the creation of firms, bankruptcy laws, administrative and taxation burden and intellectual property law. On the basis of analysis of the aforementioned factors, the project identifies the most important barriers to enterprise development through innovation, which are related to the cultural, institutional, economic, managerial/strategic and policy issues. Finally, key challenges and policy options that could lead to a more effective innovation policy are discerned.
• Determinants of innovation activities and the role of policy instruments
A firm’s propensity to innovate[1] is a complex combination of internal and external factors, which is reflected in the growing econometric and managerial literature concerned with that topic (for a partial overview see e.g. Read, 2000). Despite some occasional regularities[2], exact innovation determinants that characterise particular firms, industries, regions or countries in particular periods of time may vary and thus need to be discerned empirically – on the basis of (periodically) collected and analysed data[3]. The factors that stimulate or hinder innovative activities in particular firms can be firm or market-specific. According to Hujer and Radić (2003), market factors may include, competition intensity, market concentration, exposure to international trade, as well as demand factors like profitability and expectations regarding the future development of business. Firm specific factors can be divided into internal and external technological capabilities. Namely, sectors display different technological opportunities (which are often related to the maturity of the sector), which then affect the innovation behaviour of individual firms. Additionally, innovations can be boosted if firms co-operate with other institutions, but in order to make such co-operation beneficial, the firm must already possess adequate technological expertise, which are related to different forms of investment. External technological capabilities may be captured by variables indicating R&D-cooperation with other institutions like universities, or other firms. Internal technological capabilities may include the state of technology, the existence of an R&D department, the share of highly qualified employees and employees devoted to R&D. Other firm specific factors which are also often included in the estimation are the size of establishments measured by the number and squared number of employees, as well as industrial and regional dummies.
Policy instruments such as different forms of R&D subsidies are also an important factor that can facilitate innovation activities (cf. Criscuolo and Haskel, 2003). Formulation, implementation and modification of an effective and efficient innovation policy need to be informed by the relevant research. However, the impact of policy instruments is not unproblematic, for example due to sample selection effects. The evolution (or cancellation) of policy instruments should therefore depend upon their evaluation; where possible, it is advisable to discern their treatment effects (see below).
c) Assessing the innovation policy in Croatia: a methodology
• Outline
It is proposed that two papers are produced.The first one assesses the treatment effects of selected policy instruments and determinants of innovation activities by using a probit model that estimates their contribution to innovation propensity, controlling for the sample selection problem. The second one involves a descriptive analysis of innovation policy and its basic outcomes, complemented by relevant statistics and indices (some based on the European Innovation Scoreboard, some purposively constructed), as well as the results of the first paper. The separation of the papers is motivated by methodological and pragmatic considerations. They methodological concerns include the differences in types of data and analyses that are being undertaken, which should be reconciled with the need for focused argumentation and methodological consistency within each paper. Moreover, since the results of the microeconometric analysis will be incorporated into broader evaluation of innovation policy, it will reinforce its validity. The pragmatic concerns justifying the separation of the papers pertain to the maximisation of potential impact within the academic and policy- making communities, given that such analyses have not been undertaken in the case of Croatia so far. Namely, the analytical apparatus (i.e. the microeconometric analysis) will be applied in an academically sound way, resulting in a paper that can be considered by a relevant international economic journal. The paper addressing broader innovation policy issues would be communicated to wider social science audience and the policy-making community in Croatia and abroad, most likely through a social science / policy journal.In addition to the provision of results that can be benchmarked against other transition economies, the paper is expected to contribute (in substantive and methodological terms) to the discussion on innovation policy within the debates on transition and EU accession in Central and Eastern Europe (and Southeast Europe in particular).
• Data sources: Community Innovation Survey and others
The key source of data is being obtained through the first Community Innovation Survey (CIS) in Croatia. In this survey enterprise-level data are collected in accordance with the Oslo Manual (OECD, 1996) guidelines and the available literature on the implementation of CIS III (e.g. Kurik et al.2002; Boia et al., 2003a). The survey covers the period from 2001 to 2003. In addition to general information about the enterprise, the survey includes the data on the following aspects of innovation activities: product and process innovation, expenditures on innovation activities, intramural research and experimental development, innovation cooperation, sources for innovation, factors hampering innovation activities, innovation protection, and important strategic and organisational changes in the enterprise[4]. The survey is based on a stratified representative sample of all Croatian enterprises in relevant manufacturing and service sectors[5]. Consequently, the survey is to provide comprehensive overview of innovative activities in Croatian enterprises, which should form a basis for the formulation of more effective innovation policy. Although the survey provides a fairly comprehensive data set, it also has a number of important shortcomings that can affect the validity and usefulness of the data, which have been tackled, among others, by Archibugi and Pianta (1996), Radošević (1998) and Criscuolo and Haskel (2003)[6]. However, the shortcomings can be controlled to some extent by careful survey implementation and subsequent application of analytical techniques (for an example, see Boia et al., 2003b). Given the richness of the dataset collected within the CIS, only some of its aspects can be tackled within the proposed project. At the same time, limitations of the data and the discussion of innovation policy issues necessitate the use of aggregate data on R&D and other issues addressed by the European Innovation Scoreboard (see Appendix), available academic literature on national innovation systems, policy documents, reports etc.
• Paper on the determinants of innovation and treatment effects of policy instruments
The first paper will assess the key determinants of innovativeness of enterprises, using microeconometric evaluation methods. Thereby special attention will be spent on selected policy instruments, such as R&D subsidies, using microeconometric evaluation methods. The dependent variable of the probit model will be innovation propensity of an enterprise, measured by the decision to improve existing product/service, or to introduce an incremental or radical innovation within a specific period. As outlined above[7], explanatory variables might include investment, employees, existence of an R&D department, competition pressure, linkages with external organisations etc. The size-, regional and sector-related dummies are to be included, too. We expect that size, like in other surveys, will be one of the significant variables.
When it comes to dealing with treatment effects of policy measures, microeconometric evaluation methods will provide a starting point (cf. Hujer and Radić, 2003). One severe problem which arises in this context is due to sample selection. In recent studies it has been found that establishments which have a higher propensity to innovate are also those which are more likely to receive a public subsidy. In this case a simple mean comparison by treatment status or including such a treatment variable in a regression model will yield an upward biased estimate. Basically there are two ways to overcome this problem: Non-parametrical methods like e.g. matching-methods or parametric approaches, e.g. instrumental variable estimators. Intuitively, matching estimators try to approximate the counterfactual outcome of subsidized establishments, i.e. their outcome under no subsidy, by finding in a large group of non-subsidized those which are similar to the subsidized ones in all aspects except for the fact that they have not received a subsidy. If sample selection is solely due to observables, the difference in the outcome variables between subsidized and non-subsidized is attributable to the subsidy. The various proposed matching estimators differ in the way how the control group is constructed (for an overview see e.g. Heckman, Ichimura and Todd (1998)). The advantage of matching estimators is that they make no functional form assumptions. However, they all assume that sample selection is solely due to observable characteristics which could be a too restrictive assumption in our case. If e.g. management ability or corporate culture play an important role in both determining the innovation behaviour and the decision whether establishments will successfully apply for a public subsidy or not, matching estimators will still be plagued by a “hidden” bias. Hence, alternative estimation strategies are called for. Examples include the instrumental variable approach (see e.g. Angrist (1991)) which, however, rests on the necessity of finding valid and reliable instruments in the dataset, i.e. variables which only affect the decision to apply for a R&D-subsidy but do not affect the outcome-variable of interest.