Geographical clusters and State policy

A new Silicon Valley?
Geographical clusters and State policy
Gert-Jan Hospers, Pierre Desrochers, Frederic Sautet
The Next Silicon Valley? On the relationship between geographical clustering and public policy // International Entrepreneurship Management Journal. Vol. 5 (2009). P. 285–299.
The translation is published on the website inliberty.ruFascinated by the successes of the American Silicon Valley, various states are striving to create their own economic "cluster" — a geographically localized zone of innovative economy.

In Russia, such a project was the ambitious "innograd" Skolkovo. However, no matter how attractive the idea of a controlled innovation breakthrough is, the experience of many countries shows that the state does not have the necessary information and intuition to assess which industry or science to develop in the chosen territory. A successful "cluster" cannot be created by an official's order.

The successful experience of geographical concentration of economic activity in California inspires many states to similar efforts: they want to create their own Silicon Valley. In this paper, we conduct a critical analysis of the state "cluster policy". With the help of theoretical reasoning and concrete examples, we show that the policy of geographical concentration is a very risky business, especially when it comes to copying the "best samples". Therefore, we recommend that political leaders abandon the Silicon Valley model and start with a more modest "regional realism".

IntroductionIn a number of countries around the world, the authorities are trying to create a new Silicon Valley.

To date, Silicon Valley, the so—called area of California that includes Santa Clara and neighboring counties, is perhaps the most famous example of geographical concentration of economic activity (Saxenian 1994; O'Mara 2004). For a long time, Santa Clara County and its main cities - San Jose and Palo Alto — were famous mainly for their vineyards. However, in 1891, the former governor of California, railroad magnate Leland Stanford, founded Stanford University in the area. Under the leadership of Frederick Terman (1900-1982), this educational institution became a major training center for technical specialists and a real "incubator" of innovative companies. One of these new firms was created in 1939 by Stanford classmates Bill Hewlett and David Packard, who managed to invent a number of electronic devices. Later, semiconductors and computer chips were created in Silicon Valley, which turned into the center of microelectronics, which companies located there sell to manufacturers of electronic devices and their components from different countries.

Many government officials from many countries, delighted with this success, at various times visited the area with "working visits". One of the first such "political tourists" was the Soviet leader Khrushchev: what he saw made such an impression on him that he tried to copy the American experience by founding the famous Akademgorodok in Siberia. However, this "science city", contrary to the hopes of the first secretary of the CPSU Central Committee, did not become the Soviet Silicon Valley (Josephson 1997).

However, it was not only Khrushchev who sought to create a "valley of high technologies": many leaders and politicians from different countries tried to repeat the success of Northern California later — from Charles de Gaulle to mayors of cities (Bouwman, Hulsink 2000; O'Mara 2004; Stuart and Kargon 1996). In recent years, a number of clusters have emerged spontaneously or as part of regional initiatives for the geographical concentration of high technologies, including Silicon Alley (Manhattan, New York), Silicon Snowdrift (Minneapolis-St. Paul area), Silicon Desert (Phoenix), Silicon Mountain (Colorado Springs), "Silicon Prairie" (Champaign-Urbana), "Silicon Kingdom" (Virginia), "Silicon Hills" (Austin), "Silicon Forest" (Seattle), "Silicon Peat Bog" (Cambridge), "Silicon Gorge" (Glasgow), "Silicon Valley" (Limerick), "Medicon Valley" (Copenhagen), "Silicon Coast" (Southern Norway), "Silicon Saxony" (Saxony), "Bavarian Valley" (Bavaria), "Silicon Polder" (Netherlands), "Dommel Valley" (Eindhoven), "Silicon Kashba" (Istanbul), "Shalom Valley" (Israel), "Silicon Plateau" (Bangalore, India), "Media Valley" (Incheon, South Korea), "Billican Valley" (Arnheland, Australia), and "Telecommunication Valley" (Minas Gerais, Brazil).

Taking all this into account, we decided to analyze the question of how the successful geographical concentration of economic activity correlates with the efforts of the state undertaken for the same purpose. We wonder whether the state is capable of helping to form such "clusters", and is it possible to distinguish between "high-tech" and "low-tech" clusters? In other words, is it possible to create a second Silicon Valley with the hands of the state, or should it concentrate on the "traditional economy"? Analyzing these issues, we use theoretical concepts and concrete facts related to clusters and "cluster policy". The structure of the article is as follows: after a critical analysis of the concept of clusters, we will prove that, despite claims to the contrary, the "concentration policy" remains one of the varieties of industrial policy, which by definition implies arbitrary selectivity.

Then we will discuss the general flaws of the concentration policy, as well as the specific risks associated with the creation of high- and low-tech clusters by the hands of the state. Taking into account the fact that the state in any case always strives to encourage the concentration of production, we give examples of the successful emergence of clusters without its participation, or with minimal help from the authorities, limited to cluster marketing. All these examples show how important it is in such cases to take into account the peculiarities and realities of the area where the concentration takes place. In the final part of the article, we recommend that politicians abandon the fashionable strategy of "cloning Silicon Valley" in favor of pragmatic "regional realism".

Geographical concentration and State policy: critical analysisHarvard University professor Michael Porter, one of the leading proponents of the state policy of concentration, gives this definition of a "cluster":

It is "a group of geographically interconnected neighboring companies and associated institutions specializing in one area and united by common interests, as well as complementarity" (Porter 2000a: 254). Thus, clusters include various entities — from specialized suppliers, service companies and enterprises of adjacent industries to universities, standardization agencies, and trade associations. At the same time, it is argued that their geographical concentration facilitates the exchange of ideas and people, which, in turn, encourages and strengthens innovative activities.

Clusters: a vague conceptDespite the beautiful-sounding definition, it is not easy to identify clusters in practice, since all sectors of the economy are ultimately interdependent.

One might even say that they often exist only in the imagination of politicians and their advisers. For example, within the framework of Porter's definition, the geographical scale of a cluster can vary from a city, district, or state to several countries bordering each other. Thus, due to the extensibility of the concept itself, it is difficult for us to establish where the cluster begins and where it ends. Here's what the "cluster guru" Porter thinks about this: "The geographical scale of the cluster is related to the distance at which the increase in efficiency in the field of information, transactions, symbols, etc. is felt. Defining cluster boundaries is often a relative task: this is a creative process based on an understanding of the complementarity and interconnectedness of enterprises and institutions that are most important in terms of competition in a particular industry" (Porter 2000b: 16-17). In his opinion, there are about 60 clusters in the USA, and the OECD calls another figure — more than 300 (Martin, Sunley 2003). In one extreme variant, this concept refers to groups of industries and firms on a national scale — closely interconnected, but scattered across different regions of the country. At the other "pole" lies a different definition of a cluster — as a group of identical firms in interconnected industries located within a sharply limited spatial zone. In the broadest interpretation, even the public school system in Minnesota can be called a cluster (Rosenfeld 2001).

Martin and Sunley point out that the main explanation for this conceptual confusion is that clusters are constructs that do not have clear boundaries in terms of links between companies and sectors, information systems and geographical coverage (Martin, Sunley 2003). Because of this "elusive" nature, clusters are ideally suited as an object for various political purposes. Norton notes: "To skeptics, cluster theory seems to be a tool of regional and local politicians looking for justifications for intervention in the economy" (Norton 1999). Martin and Sunley, in addition, believe that the object of "cluster policy" is not real clusters, but something else — more easily defined and statistically "visible" industrial sectors (Martin, Sunley 2003). In practice, politicians constantly feel the need to find "clusters" in as many countries, regions or cities as possible in order not to offend certain interests of their voters. Although in theory Porter clusters do not necessarily represent highly economically specialized associations (industries), in practice almost all the "clusters" found fall under this narrow definition.

The unavoidable blurring of cluster boundaries is also explained by practical factors. Although the concentration of economic activity is usually the result of spontaneous market processes, the different capital needs of different industries result in the fact that its geographical scale varies greatly — from one street to an entire region. Moreover, we owe the very existence of cities to a unique human property — the propensity, as Adam Smith put it, "to bargain, to exchange one thing for another" (Smith [1776] 1976: 25), which in turn leads to the division of labor, including in geographical terms. Due to this tendency of people to exchange, cities have always been not completely closed and autarkic systems, but rather nodal points of trade, where individuals from different firms and networks interact on different geographical scales. As a result, even within the most advanced cluster economies like Silicon Valley, local firms tend to attach more importance to contacts with external buyers and suppliers than with their "neighbors" (Desrochers 2000).

Finally, the identification of clusters is hampered by the lack of reliable statistical information. Due to the limited data available, clusters are often identified with industries, information on which is easily found in the classification categories of national statistics. In fact, however, clusters rarely coincide with economic sectors. If the cluster elements turn out to be in different classification categories of industry and services, a problem arises: some of them, and very significant in size, can be "not seen" or incorrectly determined. Moreover, the existing industrial classification schemes are also very imperfect. As the economist Zvi Griliches (1990) noted, the "sectors" of the economy as they are presented in state statistics may in fact be nothing more than a "mirage". In particular, such a classification obscures the fact that in many cases a particular company produces a wide range of different types of products, and its employees have a variety of skills. Moreover, even the standard boundaries between industries tend to be arbitrary. For example, in the former Standard Classification System of American Industry, both "types of products" and "production processes" were used as criteria for assigning enterprises to a particular category, but at the same time such important categories as the manufacture of plastics and electronics were simply absent in it (Desrochers 2001).

Why is the "cluster policy" inevitably characterized by selectivity"Cluster policy" refers to any efforts by the state to create and support clusters in a certain area.

It is often regarded as something more modest and less ambitious than traditional industrial policy aimed at "achieving by specific industries (and the firms they consist of) the results that the state considers effective for the economy as a whole" (Chang 1994: 60). The tools used to influence a particular industry include import tariffs, subsidies to sectors experiencing difficulties and emerging industries, as well as government measures to stimulate investment in R&D. Porter states: "The ideological basis of cluster theory and industrial policy, as well as their practical significance for the activities of the state, fundamentally differ" (Porter 2000a: 27). In his opinion, the scope and mechanisms of cluster policy are broader than those of industrial policy. In fact, Porter believes that the goal of concentration policy is to promote the "competitiveness" of a country or region (Ibid.). In his opinion, areas where enterprises are vertically and horizontally grouped into clusters have greater competitiveness. The concept of "competitiveness" refers to the quality of the business atmosphere in the region, as well as to "framework" conditions such as the availability of raw materials, skilled labor and especially strong clusters. In this regard, cluster policy should be aimed at "removing obstacles, easing restrictions and overcoming inefficiencies that hinder productivity growth and innovation within the cluster" (Ibid., 26).

Ideally, the state implementing cluster policy should shift its attention from planning and subsidizing specific industries to encouraging the emergence and functioning of clusters in the economy as a whole. In the scientific literature, this change in political orientations is regarded as a transition from private to general, from direct state intervention to indirect and from vertical to horizontal policy (see, for example: Chang 1994; McDonald, Dearden 2005). The course of concentration can be considered as part of a new "unorthodox" approach, in which the essential elements of general economic policy are revised from a territorial point of view (Storper, Scott 1995). In practice, however, the "cluster" approach is difficult to isolate from other methods of economic policy. At the same time, initiatives of a "cluster" nature in policy documents and statistics can be found in several categories — such as industrial policy, innovation promotion policy and regional development policy. For example, in the ministries of economy, the same departments that were previously engaged in industrial policy are mainly responsible for the policy of geographical concentration, and the departments for regional affairs transform "network" regional programs into "cluster" initiatives.

At first glance, the framework policy of maintaining clusters in order to increase competitiveness really seems to be more comprehensive and "market-based" than traditional government intervention in the form of direct assistance to "national champions" or "appointment of winners in advance" (Chang 1994). However, on reflection, it becomes obvious that cluster policy also provides for a selective approach. Firstly, the concept of "competitiveness" itself contains an essential element of selectivity (Reinert 1995). In order to take measures to increase the competitiveness of a territory in relation to others, government officials must decide which types of economic activities should be supported and which should be left to the will of market forces. Secondly, even if the goals of the concentration policy are general in nature, the means to achieve them are indirectly related to the creation of privileges for certain types of activities. In particular, public investment in the scientific base of the relevant territory or subsidies for R&D cannot be provided to all clusters: partiality towards certain industries is inevitable (Cowling et al. 1999). For example, a biotech cluster is more likely to receive state support within the framework of a policy of encouraging innovation than those that represent "old" sectors of the economy (say, the coal industry) — because there are opportunities for radical renewal are not so obvious. Thus, the authorities — whether they focus on clusters or something else - cannot avoid selective planning.

In general, Porter's attempts to distinguish between concentration policy and industrial policy look unconvincing. Since the same problems arise in connection with cluster theory and industrial policy regarding selectivity and practical implementation, it is quite difficult to talk about the fundamental differences between these two concepts. Porter, it would seem, managed to clearly distinguish them, but in practice cluster policy is also associated with "appointing winners" or "supporting losers" (Norton 1999), and it cannot be called a radical antithesis of previous methods. Indeed, when developing a concentration policy, the authorities need to make clear decisions: it should be either "offensive", providing for the stimulation of high-tech clusters (information, bio- and nanotechnology), or "defensive", aimed at preserving traditional industries (for example, the automotive, textile and machine tool industries). In short, their goal is to create either a new Silicon Valley or a new "rust belt" (Hospers 2004a). We will see that such planning already has a long history, and it is a history of failures. Therefore, managers responsible for cluster policy should carefully avoid past mistakes.

Risks associated with attempts by the state to create geographical clustersIn practice, politicians often identify clusters with industries and act accordingly.

Moreover, as we have seen, the policy of concentration is impossible without arbitrary selection, and therefore resembles industrial policy. In this section, we will talk about the risks that cluster policy is fraught with. To begin with, let's focus on the general difficulties associated with selectivity. Then we will analyze the specific problems that the state's "selection" of high-tech or low-tech clusters is fraught with.

Planning and information asymmetryGiven that the state cannot equally support all clusters, it needs to decide which clusters should be given special attention and which should not.

To justify the choice in favor of certain clusters, politicians mainly use arguments from the arsenal of economic science. However, these arguments are quite amenable to criticism; often they are not so much scientific as political in nature, and ignore the theoretical and practical arguments against selective planning. Regarding the "scientific validity" of cluster policy, Bass notes: as a rule, "the study of industries in order to select candidates for state support is carried out on the basis of insufficient or unsuitable data, sociological methods suffering from deep flaws, and simplified mathematical models. On the basis of this, guidelines are developed, which, therefore, are of a dubious nature" (Buss 1999: 343). According to Bass, strategies related to selection, including the selection of clusters, are practiced not because of their high scientific quality, but because of their attractiveness to certain categories of the electorate. Impressive analytical works (such as the research of Porter and his team) can be "produced" on request, to substantiate inherently political proposals. And since they seem "scientific", a "herd effect" quickly appears among political consultants. When politicians in some regions begin to develop similar plans, others find it necessary to follow their example.

Economist Joseph Schumpeter has long since revealed the connection between the political process and the economic course of the state: "Political maturity cannot be achieved without understanding that the measures taken in the economy are also politics. This truth is especially often overlooked by economists" (Schumpeter 1950: 8). In addition to Schumpeter's cynical conclusion, there are also justified reasons from the point of view of economics why politicians are unable to carry out "cluster selection" more effectively than market actors. The theory of public choice says: due to the constant information asymmetry and strategically determined behavior of politicians and officials, "state failures" occur in our lives no less often than "market failures" (Wolf 1990). Among other things, government officials tend to have a poor understanding of the dynamics of business life and are too far away from the economic process to identify those areas where favorable opportunities really open up. After all, there are, by definition, fundamental differences between the public sphere (politics) and the private sphere (commerce).

Since both spheres have their own moral basis and represent different "survival systems", interference of one system in the activities of the other should be avoided whenever possible (Jacobs 1992). There is a particularly high probability of a serious information asymmetry between entrepreneurs and political leaders in matters related to cluster policy. The reason is the "implicitness" of the formation of spatial clusters. In fact, geographical agglomeration is closely related to the emergence and transmission of "implied" knowledge based on "special conditions of time and place" (Hayek 1948). Within the cluster, the most important thing is not so much explicit, formalized and easy-to-transmit information, as non-standard knowledge embodied in human capital and acquired through experience. Such knowledge is almost impossible to systematize: they can only be bought and sold when their carriers — people — move to another job. This is the total experience that firms in the cluster acquire by hiring local "veterans". Such knowledge is possessed only by individuals who have worked for a certain time in one industry or firm. This "implicit" nature of concentration is one of the explanations for the special "industrial atmosphere" that Marshall drew attention to at the end of the XIX century, studying the successful formation of "clusters" of that time like the "industrial area" around Sheffield (Marshall [1890] 1947: 225). Marshall notes: there is something special "in the air", thanks to which "it is worth one person to express a new idea, others pick it up and supplement it with their own proposals; thus, it becomes a source of other new ideas."

MacDonald points out that government officials cannot perceive these flows of implicit knowledge, which are the source of wealth of thriving clusters (MacDonald 1992). They don't have the skills, contacts, or experience of the cluster members they are assigned to deal with. It is impossible to predict in advance what steps will need to be taken in relation to the cluster, since it is impossible to determine in advance which activities within its framework will require support — it depends on the specific circumstances of the time and place. Moreover, in most cases, economic policy does not even affect knowledge about the current realities of a particular market or its possible future state. The history of the concentration of microelectronics production in Silicon Valley shows that the birth, life and death of clusters is an element of a spontaneous order based on new entrepreneurial ideas and the sum of explicit and implicit knowledge accumulated over decades. Of course, clusters are important for the efficient functioning of the economy. But government officials should understand that they are often formed spontaneously and in the process of their emergence, implicit knowledge of the "local scale" plays a fundamental role. It is impossible to predict in which areas clusters may arise in the future. Ultimately, such a concentration of production is the result of entrepreneurial activity, and its driving force is the desire to make a profit. The state is simply unable to replace the market in the process of cluster formation (Sautet 2002b).

"Designated" winners: "trendy" high-tech clustersThe governments of most countries, inspired by the phenomenon of Silicon Valley, are trying to increase the competitiveness of the economy by creating high-tech clusters.

In general, high technologies, for example, information, communication, bio- and nanotechnology are considered "modern" and innovative, and therefore the authorities are trying to stimulate their development (Drucker 1985). However, the policy of concentration of high-tech industries is fraught with at least three dangers.

Firstly, as we have demonstrated in detail above, there is no good reason to believe that the state leadership has more information than entrepreneurs to assess the economic potential of specific economic projects (including clusters). Due to the initial uncertainty about the prospects of new technologies when creating high-tech clusters, the probability of a "failure of the state" is especially high. According to Schmuckler, almost all the examples of innovation activity studied by him were stimulated not by state support for scientific research, but by the awareness of the need to solve the problem of high costs or the desire to take advantage of the opportunity to make a profit (Schmookler 1966: 199). According to Miller and Cote, this is one of the main reasons why projects to create "innovation centers" and other "incubators of ideas" at technoparks in the USA and Canada in the 1970s and 1980s invariably failed (Miller, Cote 1985). The risks associated with the strategy of "appointing winners in advance" were also manifested during the implementation of the high-tech development policy in France in the 1980s. After five years of subsidizing the production of microelectronics, Paris had to admit that it had bet on the wrong horse. One of the reasons for the failure of this venture of the French government was the lack of a "commercial vein" among the political elite — its only goal was to make France a world leader in the field of microelectronics. If you need a very recent example, indicating that the state does not understand the trends of scientific and technological progress, it is enough to recall what a fuss there was in the world around information technology. Of course, the impact of these technologies is very significant, but they, contrary to the hopes of many governments, have not become the basis of the "new economy" (Clarke 2001).

Moreover, the chances of additional profit for regions where high—tech clusters arise may be very limited - and not only because much fewer jobs are usually created in high-tech industries than in low-tech or non-technological ones (Drucker 1985). The main thing here is something else: during the global innovation race, most regions rely on the same types of activities. As a rule, state authorities tend to "keep up with others", and therefore, in almost all cases, support is provided to the development of information, bio- and nanotechnology clusters. Today, they are trying to create "silicon valleys" all over Europe (for examples, see in the introduction). However, as it appears from the scientific literature on the organization of industry, a real competitive advantage can be obtained not by imitating rivals, but by entering the market with unique products (Martin 2001). Copying Silicon Valley, other regions do not give an answer to the question of how their own clusters differ: on the contrary, they only strengthen the "pioneer advantage", that is, the California high-tech center itself. In addition, if we think in supranational categories (on the scale of the whole of Europe or North America), the current "cluster race" turns into a "duplication" of investments in the same high technologies. Such a "herd effect" can lead to the appearance of excess capacity, the formation of "soap bubbles" and a crisis during which only the strongest will survive (Lux 1995). Similarly, by investing in the same technologies and copying the "best samples", regions undermine their potential competitive advantage, and it should not be surprising if all this ends in a painful "shake-up".

Finally, while supporting high-tech clusters, the authorities often do not ask the question: are there prerequisites for their emergence in the relevant region? Here, differences in the starting situation, the structure of the economy and institutional features are of great importance. Representatives of the "evolutionary" trend in economics have demonstrated that innovations are often generated by random factors or a unique socio-economic situation (Witt 1993). Thus, the methods that worked in one region are not always suitable for another territory. For example, a high-tech cluster can hardly be successfully created in an area that does not have the "absorbency" in terms of new technologies (Cooke 2002). If there is no such "host system" in the region, the concentration policy is fraught with serious risks. Castells and Hall provide convincing evidence that the costs of creating clusters "from scratch" are very high, and their formation is a very long process (Castells, Hall 1994). An example of the unsuccessful creation of a high-tech cluster by the hands of the state can be the Russian Akademgorodok. In the 1950s, this Siberian "science city" was built on the model of Silicon Valley literally "from scratch". Since then, he has been dragging out a miserable existence for many decades. Other examples showing the importance of compatibility with local conditions when creating clusters are associated with Southern Italy and the Ruhr (Hospers 2004a). Here, the local environment simply rejected ambitious state projects implemented in the 1960s and 1970s. The industrial complexes in Sardinia and the high-tech sector in the Ruhr turned out to be "dummies" and turned into "Egyptian pyramids".

Support for "losers": traditional low-tech clustersHowever, the state policy of concentration of economic activity provides not only for the support of high technologies.

Many regions have inherited from the past a "traditional economy" of a low-tech or non-technological nature (Hayter 1997). Due to intense competition and declining demand, the "old" centers of the textile, food, automotive, metallurgy and shipbuilding industries found themselves in a difficult situation. Although these "national champions" have been restructured in a number of cases over the past decades, most of them still receive assistance from the State within the framework of "cluster" or regional policy (Todtling, Trippl 2004). Is such a policy of supporting low-tech industries an effective alternative to the "fashionable" trend of copying Silicon Valley?

Unlike many high-tech projects, such clusters are at least rooted in the local environment and give jobs to a large number of people (Fingleton 1999). From the perspective of the now popular "new economic geography", it can be said that the increase in profits due to the scale of production in these traditional economic activities lays the foundations for a future competitive advantage — albeit in the long term (Puga 2002). Thus, such clusters have a chance to prove their economic feasibility. Examples include coal mining and metallurgical production in Wallonia (Belgium), the forestry industry in Scandinavia and Canada, the automotive industry in southern Germany, watch factories in the Swiss "valley of precision", the textile industry in northern Italy, the production of snow removal equipment in Finland, irrigation equipment in Israel and winemaking in California. However, the nature of state support for such clusters in most cases does not stand up to criticism — the authorities "help losers".

Firstly, the state policy of supporting low-tech clusters usually pursues several different goals at once, which already excludes the optimal result. The desire of the authorities to preserve the "national champions" is often chaotically intertwined with restructuring, employment and considerations of national industrial policy. Scientists do not tire of emphasizing that such a variety of goals within the framework of one political direction is dangerous, since they can come into conflict with each other and thereby hinder the development of a clear strategy (Dunn 2003). In fact, here we are dealing with the famous principle of "matching tools and tasks" deduced by Tinbergen: in the economy, using one common tool, it is impossible to achieve several different goals, for example, ensuring full employment and sustainable growth. On the contrary, each goal requires its own special method (Tinbergen 1952). The clearest example of the neglect of this principle can be considered the policy of support for shipbuilding, which the Dutch government has been conducting since the 1970s. Time after time, short-term considerations like maintaining employment in the north of the country (in Groningen and Friesland), as well as the factor of national pride, prevail over reasonable economic arguments in favor of reducing the industry. The political leadership of Britain has made the same mistakes with regard to the domestic automotive industry since the 1970s. The desire to simultaneously achieve goals of a social, structural, patriotic and economic nature has failed: this may explain the decline of the British automotive industry — from the bankruptcy of British Leyland to the very recent collapse of Rover.

In addition, the policy of supporting low-tech clusters in the form in which it is implemented, most often "treats the symptoms, not the disease." Often these programs provide subsidies to industrial companies that find themselves in a difficult financial situation. In theory, temporary support for traditional industrial clusters to allow them to "come to their senses" is justified (Todtling, Trippl 2004). The problem is that it is easier to start subsidizing than to stop it. Worse, there is the possibility of "drug addiction" from subsidies, when enterprises can no longer exist without state aid (Howitt 1996). After all, subsidizing traditional industries often does not contribute to the restructuring of companies, but on the contrary, reinforces the inefficiency of their activities inherited from the past. Representatives of the institutional school in economic geography note that traditional industrial regions are particularly vulnerable to such a danger (Fuchs, Shapira 2005). This is due to the phenomenon of "inertia" and idiosyncrasy on "foreign" inventions, i.e. the tendency characteristic of regions that flourished in the past to cling to traditional models instead of keeping up with the changed economic realities. Typical examples in this regard are Wallonia and the Ruhr in the 1960s. For a long time, the "bond" between local industrialists and politicians led to the artificial maintenance of employment in the coal and metallurgical industries, which delayed the reorientation of regions to new types of activities. In short, the policy of state support for low-tech clusters may hinder the restructuring necessary to bring production in line with demand. As a result, the region may miss the opportunity to fit into new market conditions.

If you can't help, at least don't hurtSo far, we have criticized the "cluster policy", noting that in fact it is equivalent to industrial policy — with all the associated risks.

But while recognizing that the state's capabilities in terms of the successful implementation of industrial policy are very limited, someone may nevertheless argue that it is able to play a certain role - to contribute to the emergence of clusters without creating them directly. The issue of incentives as a policy tool is not new, but recently it has become particularly relevant — primarily in Anglo-Saxon countries. The Ministry of Industry of New Zealand, the Ministry of Trade and Industry of Great Britain and the Ministry of Industry of Canada, for example, consider support for the formation of clusters one of the main functions of the state, equivalent to, say, infrastructure development. Maybe this approach avoids the dangers associated with "appointing winners" and "supporting losers"?

In practice, the creation by the state of conditions for the emergence of clusters can lead to the same result as direct intervention in economic processes within the framework of the "concentration policy". In particular, there may be similar problems with information and incentives. The Austrian economist Kirzner has repeatedly stressed that the main danger associated with state economic policy (he called it "regulation") is that it stifles innovative entrepreneurship (Kirzner 1979). In his opinion, there are several main reasons for this:

1. Ignorance of regulatory authorities about alternative options.

2. The inability of regulatory authorities (due to lack of motivation — profit) to identify opportunities for improving coordination.

3. The potential stifling effect of regulation on the innovation process.

4. The probability that regulation may direct market development to a path undesirable from the point of view of consumers (Kirzner 1979; Sautet 2002a).

By stifling entrepreneurial activity, regulation weakens the market's ability to generate knowledge that can improve the coordination of individual actors' plans. In other words, the fundamental problem with the "market-oriented" state policy (including the stimulation of certain market processes) is that it reduces the range of coordination functions of the market mechanism.

As we have already emphasized above, the policy of stimulating the emergence of clusters often boils down to "appointing winners" and "supporting losers". This is due to the fact that the state is unable to generate the knowledge necessary for the successful functioning of clusters. And there is no reason to believe that in the institutional conditions associated with the active actions of the state to stimulate the emergence of clusters, such knowledge may appear. Thus, even if there were people working in the state apparatus who care exclusively about the common good, they would still not be able to successfully pursue a "cluster policy". The limited effectiveness of the active actions of the state is due not so much to human nature as to the lack of information necessary for the successful solution of the tasks set (Sautet 2002b). From this point of view, the state policy of stimulating the emergence of clusters is just another variation on the theme of industrial policy. This is a way of "disguised" state intervention in the economy. The reason is that such a policy can again take the form of subsidies — by definition biased and also unnecessary. After all, if subsidies are provided to clusters that would have been successful without them, there is no need for this assistance. If firms are subsidized that would otherwise face bankruptcy, this is tantamount to interference in the process of market "natural selection" — the artificial preservation of enterprises that otherwise would simply not exist.

The state policy of stimulating clusters can fall into the same trap of "appointing winners" and "supporting losers". At any given time and in any place, the amount of available resources is limited. Therefore, the agencies entrusted with this stimulation, one way or another have to choose which clusters to provide assistance to, and which ones not. Since, as Kirzner notes, there is no reason to believe that the state has incentives and knowledge that allow it to successfully carry out such a selection, stimulation is no different from other varieties of "cluster policy" (Kirzner 1979). At any moment and in any place, the optimal structure of industry can arise only through the medium of market mechanisms, by trial and error. In general, the State should completely avoid interfering in the coordination functions of the market and not limit the forms that the process of change takes. In the conditions of a "living economy" (to use the term of Mises), the birth and death of cities, industries, industrial centers are inevitable, and the state should not slow down this process. This fact should serve as a warning to the political leadership in general and officials in charge of "cluster policy" in particular. In a word, the famous Hippocratic principle "Do no harm!" (cit. by: Salacuse 1994).

What should be the role of the state?If creating or encouraging the creation of clusters by the hands of the state is fraught with risks, is a "cluster policy" appropriate at all?

To answer this question, let's briefly look at some examples of successful cluster creation and see what contribution the state has made to this process. Table 1 provides a short list of "success stories" in Europe (for a more complete list, see: Hospers 2004b).

Table 1. Examples of successful cluster creation in Europe due to a combination of traditions and modern global trends

RegionTraditional industry	Global trend	A new combination	Jura d'Arc
	Watchmaking	Marketing and Lifestyle	Swatch Watches
Emilia-Romagna	Textile industry	High production technologies	Fashionable clothes
Baden-Württemberg	Machine tool construction	Digital media	Multimedia equipment
Jutland	Furniture manufacturing	Quality and lifestyle	Designer furniture
Manchester	Heavy industry	Pop music/pop art	Entertainment Industry
North of Pas-de-Calais	Clothing production	Striving for convenience	Mail orders
Ruhr	Heavy industry	"Economics of experience"	"Industrial culture"
Dunakanyar	Tourism	Population ageing and rising living standards	Health resorts
Krakow region	Construction and finishing	Demand for the maintenance of historical heritage	Restoration services

As can be seen from this list, in some cases we are talking about the revival of traditional economic activities through the introduction of high technologies in the field of design, production and marketing. This strategy resulted in "new combinations" in the Swiss watch industry, the textile industry in Italy and furniture production in Denmark. The origin of the Manchester cluster in the UK, specializing in pop music and other forms of art, and the multimedia cluster in Baden-Württemberg (Germany) is connected with the industries that previously existed there, which allowed us to acquire skills in handling modern materials and proved useful in new fields of activity. Other regions have taken advantage of the growing demand for consumer services. In the Northern Pas-de-Calais (France), several textile factories have been transformed into mail-order clothing factories, and in the Ruhr region (Germany), closed mines and factories are used as tourist attractions (for lovers of "industrial culture"). Examples of such a combination of "tradition and modernity" in Central and Eastern Europe include modern health resorts in Hungary and high-tech restoration services of the Polish construction cluster.

Three things are very important to explain the effective functioning of these clusters. Firstly, the examples given indicate that successful clusters are almost always based on the economic structures existing in the region. The age-old traditions of, say, the watch industry in Switzerland, coal mining and metallurgy in the Ruhr and recreation at healing springs in Hungary laid the foundation for modern clusters in these regions. Obviously, the path of the future development of the regions depends on where they started (Hassink 1997). In a word, let this observation be trivial, but the fact remains that the economic prospects of the region are inevitably connected to its past in one way or another. Thus, the examples given do not concern "the best samples", but rather "unique models", demonstrating only that the competitiveness of a region depends on its inherent features. Secondly, in order for the development of existing potential not to turn into "support for losers", traditions must be combined with modern trends. In fact, the success of clusters — to paraphrase Schumpeter's definition of innovation (Schumpeter 1912) — is always the result of a "new combination" of traditions and global trends. As the table shows, such interactive interaction at the global and local levels, creating opportunities for growth, can arise as a result of the "addition" of existing economic activities in the region, modernization of traditional industries for inclusion in the "new economy" or the use of knowledge and skills accumulated in the past to solve new problems caused by structural changes across the entire economy. Thirdly, evaluating the examples given, it is impossible not to be amazed at how insignificant a contribution the state has made to the success of these clusters. All of them arose spontaneously, and if the state played a role in their development, it was already at subsequent stages.

Let's take a closer look at what the state did in connection with the formation of clusters listed in the table (Hospers 2004b). In some regions, it almost did not participate in this. This was the case in Manchester, Pas-de-Calais, Dunakanyar and the Krakow region. Of course, today the authorities of these regions advertise the clusters available there to attract tourists and investments, but they do it after the fact — after their spontaneous formation. In Emilia-Romagna, Baden-Württemberg and Jutland, the state helped to create business support and technology implementation centers that provided "real services" to clusters (for example, technical consultations and the organization of networking events). The state did not interfere in the activities of the clusters, but only provided information and contacts at the request of the business community. Of course, this can be considered as a kind of "stimulation", but due to the specific nature of the state's participation, which is not associated with intervention and was carried out after the emergence of clusters, it did not play a negative role. In other cases, the state was limited only to marketing clusters. In the Swiss Jura, Baden-Württemberg and the Ruhr, it was the local business circles that decided to join forces and create clusters for the production of watches, multimedia technologies and "industrial tourism". When it became obvious that this activity was yielding successful results, the state took over the "branding" of clusters on an international scale. Thus, the respective regions were presented as "places that would be worth visiting" to parties (investors, tourists) interested in the types of activities that these clusters are engaged in. This state "advertising" does not harm market processes as long as its purpose is to attract the attention of new investors and customers who are able to participate in market activities (Kotler et al. 1993; Rainisto 2003). The saying "If you're good, don't be shy about talking about it" also applies to clusters: their task is to act effectively, and the state can take over the dissemination of information about clusters that have passed the market test.

Conclusion: about the benefits of "regional realism"Inspired by the success of Silicon Valley, many countries of the world are pursuing their own "cluster policy" in an attempt to copy it.

Many see this as something less pretentious than traditional "selective" economic policy. According to Porter, such stimulation of the concentration of production — in contrast to industrial policy — is a horizontal, market-oriented approach (Porter 2000a). In this article, we have criticized this point of view — for several reasons. Firstly, due to the vagueness of the very concept of "cluster", these entities are often identified with industries and are perceived by the authorities as such. Moreover, cluster policy (like traditional industrial policy) is characterized by arbitrary selectivity: the authorities choose which clusters they will help and which they will not. In general, such steps are fraught with risks due to the serious information asymmetry that arises between the political leadership and entrepreneurs. There are especially many problems when it comes to supporting high-tech and low-tech clusters. The first strategy resembles industrial policy with its "appointment of winners" and the accompanying "traps", and the second is a dangerous course to "support losers".

At first glance, the state policy of indirectly stimulating the emergence of clusters looks more preferable and desirable, but we argue that in this case, the authorities face the problem of lack of knowledge about market processes. Thus, the best advice to officials responsible for cluster policy will be the Hippocratic covenant "Do no harm!". Indeed, analyzing examples of the creation of successful clusters, we see how little the state can help this process. In our opinion, the authorities can play a role here only at the last stage - to disseminate information about the achievements of clusters after they spontaneously arose in market conditions. When implementing this "cluster marketing", it is important to keep in mind: it is also necessary to talk about the realities and features of the respective regions. After all, competition is not copying someone else's samples, but creating something unique. Obviously, attempts to build "their own" Silicon Valley do not fall under this definition. If States want to contribute to the emergence of clusters, they should approach the matter on the principle of "Do no harm!". With this in mind, the authorities do not have a large choice of political tools: it may be enough not to interfere with the natural process and reduce taxes (as well as simplify regulations when creating new companies). Thus, the authorities, inspired by the concept of clusters, should show modesty and be guided by "regional realism", and not dreams of a new Silicon Valley.

LiteratureBouwman, H., & Hulsink, W. (2000).

• Silicon Valley in de Polder: ICT-clusters in de Lage Landen. Utrecht: Lemma.
• Buss, T. F. (1999). The case against targeted industry strategies // Economic Development Quarterly, 13, 339–356.
• Castells, M., & Hall, P. (1994). Technopoles of the world: The making of twenty-first-century industrial complexes. London: Routledge.
• Chang, H. J. (1994). The political economy of industrial policy. London: St. Martin’s.
• Clarke, R. (2001). The New Economy beyond the hype: The OECD Growth Project. Paris: OECD.
• Cooke, P. (2002). Knowledge economies: Clusters, learning and cooperative advantage. London: Routledge.
• Cowling, K., Oughton, C., & Sugden, R. (1999). A reorientation of industrial policy: Horizontal policies and targeting. In K. Cowling (Ed.) Industrial policy in Europe: Theoretical perspectives and practical proposals. London: Routledge.
• Desrochers, P. (2000). Geographical proximity and the transmission of tacit knowledge // Review of Austrian Economics, 14, 63–83.
• Desrochers, P. (2001). Local diversity, human creativity and technological innovation // Growth and Change, 32, 369–394.
• Drucker, P. (1985). Innovation and entrepreneurship. London: William Heinemann.
• Dunn, W. N. (2003). Public policy analysis: An introduction / 3rd ed. New York: Prentice Hall.
• Fingleton, E. (1999). In praise of hard industries. Boston, MA: Houghton-Mifflin.
• Fuchs, G., & Shapira, P. (2005). Rethinking regional innovation and change: Path dependency or regional breakthrough? New York: Springer.
• Griliches, Z. (1990). Patent statistics as economic indicators // Journal of Economic Literature, 28, 1661–1707.
• Hasssink, R. (1997). What distinguishes ‘good’ from ‘bad’ industrial agglomerations? // Erdkunde, 51, 2–11.
• Hayek, F. (1948). Individualism and economic order. Chicago: University of Chicago Press.
• Hayter, R. (1997). The dynamics of industrial location. Chichester: Wiley.
• Hospers, G. J. (2004a). Restructuring Europe’s Rustbelt: structural change in the German Ruhrgebiet // Intereconomics: Review of European Economic Policy, 39, 147–156.
• Hospers, G. J. (2004b). Regional economic change in Europe: A neo-Schumpeterian vision. Münster; London: LIT.
• Howitt, P. (1996). The implications of knowledge-based growth for micro-economic policies. Calgary: The University of Calgary Press.
• Jacobs, J. (1992). Systems of survival: A dialogue on the moral ffoundations of commerce and politics. New York: Vintage.
• Josephson, P. (1997). New Atlantis revisited: Akademgorodok, the Siberian City of Science. Princeton: Princeton University Press.
• Kirzner, I. M. (1979). The perils of regulation: a market-process approach. Reprinted in I. M. Kirzner 1985. Discovery and the capitalist process. Chicago: University of Chicago Press.
• Kotler, P., Haider, D., & Rein, I. (1993). Marketing places. New York: The Free.
• Lux, T. (1995). Herd behaviour, bubbles and crashes // The Economic Journal, 105, 881–896.
• MacDonald, S. (1992). Formal collaboration and informal information flow // International Journal of Technology Management, 7, 49–60.
• Marshall, A. ([1890] 1947). Principles of economics. London: MacMillan.
• Martin, S. (2001). Industrial organization: A European perspective. Oxford: Oxford University Press.
• Martin, R., & Sunley, P. (2003). Deconstructing clusters: Chaotic concept or policy panacea? // Journal of Economic Geography, 3, 5–35.
• McDonald, F., & Dearden, S. (2005). European economic integration / 5th ed. Essex: Pearson Education.
• Miller, R., & Côté, M. (1985). Growing the next Silicon Valley // Harvard Business Review, 114–123, July–August.
• Norton, R. (1999). Cluster theories: Spatial externalities // The web book of regional science / Ed. by R. W. Jackson. West Virginia University: Regional Research Institute.
• O’mara, M. (2004). Cities of knowledge: Cold war science and the search for the next Silicon Valley. Princeton: Princeton University Press.
• Porter, M. (2000a). Location, competition, and economic development: Local clusters in a global economy // Economic Development Quarterly, 14, 15–34.
• Porter, M. (2000b). Location, clusters and company strategy // The Oxford handbook of economic geography / Ed. by G. L. Clark, M. S. Gertler, & M. F. Feldman. New York: Oxford University Press.
• Puga, D. (2002). European regional policy in light of new location theories // Journal of Economic Geography, 2, 372–406.
• Rainisto, S. (2003). Success factors of place marketing: A study of place marketing practices in Northen Europe and the United States. Helsinki: Helsinki University.
• Reinert, E. S. (1995). Competitiveness and its predecessors: A 500-year cross-national perspective // Structural Change and Economic Dynamics, 6, 23–42.
• Rosenfeld, S. (2001). Backing into clusters: Retrofitting public policies. Paper presented at the Integration Pressures: Lessons from Around the World-John F. Kennedy School Symposium at Harvard University, March 29–30. Salacuse, J. W. (1994). The art of advice. New York: Crown.
• Sautet, F. (2002a). Economic transformation, the pretence of knowledge and the process of entrepreneurial competition. New Zealand Treasury Discussion Paper.
• Sautet, F. (2002b). Kirznerian economics: Some policy implications and issues // Journal des Economistes et des Etudes Humaines, 1, 131–151.
• Saxenian, A. L. (1994). Regional advantage: Culture and competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press.
• Schmookler, J. (1966). Invention and growth. Cambridge: Harvard University Press.
• Schumpeter, J. A. (1912). Die Theorie der Wirtschaftlichen Entwicklung. Leipzig: Duncker and Humblot.
• Schumpeter, J. A. (1950). American institutions and economic progress. Reprinted in Swedberg, R. 1991. Joseph A.
• Schumpeter: The economics and sociology of capitalism. Princeton: Princeton University Press.
• Smith, A. ([1776] 1976). An inquiry into the nature and causes of the wealth of nations. Oxford: Clarendon Press, (textual editor W.B. Todd).
• Storper, M., & Scott, A. J. (1995). The wealth of regions // Futures, 27, 505–526.
• Stuart, W. L., & Kargon, R. (1996). Selling Silicon Valley: Frederick Terman’s model for regional advantage // Business History Review, 70, 435–472.
• Tinbergen, J. (1952). On the theory of economic policy. Amsterdam: North-Holland.
• Tödtling, F., & Trippl, M. (2004). Like Phoenix from the ashes? The renewal of clusters in old industrial areas // Urban Studies, 41, 1175–1195.
• Witt, U. (1993). Evolutionary economics. Cheltenham: Edward Elgar.
• Wolf, C. (1990). Markets or governments: Choosing between imperfect alternatives. Cambridge, MA: The MIT Press.

Portal "Eternal youth" http://vechnayamolodost.ru24.03.2011