A brief history of clusters…

Posted on 29/06/2009


Clusters are essentially a social fact.

They appeared in the history of humankind because of two elements: technical innovations and the bigger size of human groups. Technical innovations generated a need for specialization. When groups of humans began to develop more and more sophisticated tasks, appeared at the same time the first specializations of people in a group. We can date this age of technological development at the Neolithic, around 9500 BC. With the development of the size of groups and the need to master more and more specific techniques, came the specialization of some members of the group on specific activities, such as farming, raise of domesticated animals and of course use of more and more sophisticated tools in bronze and later iron. As size of groups increased, after having only one or two members for example mastering the craft of weapons and tools or the transformation of skins into leather, there were gradually 10 or 20 or a 100. Because of human social behaviors mostly, people in the same activity tend to gather together. This is the first stone in the creation of clusters. Being together allows for an exchange of issues and best practices and ultimately generate not only innovation but also a faster diffusion of these innovations. This need of being physically together led to the creation of dedicated areas in cities. Each city has quarters dedicated to specific groups of craftsmen. This situation evolved gradually over time and was more or less stabilized at the end of the Middle Age (16th century).

Again, having specific activities in the same area allowed for several benefits: . a faster exchange of information . a mutual support between the members of the group

 . a faster solving of issues

. a faster diffusion of innovation

. an easier and cheaper access to raw supply

. a better access to skilled manpower .

The clusters of small businesses that we describe here were a self-creation of the different craftsmen as they were seeing a clear benefit in evolving in a system of coopetition (competition-cooperation). There are in most European capitals remains of this situation in the names of streets linked to the type of business which was active in the area: Dyer Street or “leather finishing plants” Quay in Paris. It is interesting to notice that if at that time, most of the clusters were a self-creation, some of them were the results of rules and regulations, precluding what would happen in the 20th century but for other reasons. For comfort reasons for example, because the activity was smelly and dirty, the businesses of transforming leather were asked to be located in a specific area. Or it could be for safety reasons, the best example being the glass makers in Venice who were asked to move in 1291 to move from Venice to Murano because of the risk of fires in the city. Today, glass makers are still in Murano and constitute a cluster of SMEs, as do for example the leather transformation businesses in large cities in Marrakech, and still for the same reasons. On top of the existence of clusters of craftsmen in every (large) city, this period saw the creation of clusters at the regional, country and then continental level. One city or one region would become the specialist of one activity and have a competitive advantage at the regional, national or European level.

Why is that so ? Again, we find technical innovations leading to the need to specialize activities for the benefits of all. We have to add to that factor a second one: means of transportation. As it became “easier” along the Middle Age and then the Renaissance to trade at the level of a region or a continent (and even cross continents), local groups of craftsmen could gain a competitive advantage over groups in other cities or countries and become competitive clusters at a larger level. This situation remained the same in Europe until the beginning of the 20th century and faded regularly to almost disappear now. What made these clusters of craftsmen disappear? Again, it is mostly due to the emergence of faster and more reliable communication means during the 20th century. Trains, safer roads, telephones, cars, and later cargos and planes, all these innovations reduced the need to be close to each other to benefit from a better supply of raw materials and skilled manpower and allowed to develop a competitive advantage at the international level. From self-emerging clusters to government supported clusters As we have seen previously, clusters along history were more or less a self creation of the SMEs in a particular industry. This situation almost disappeared in the 20th century to be replaced by a situation where clusters are created and supported by governments.

Where does this shift come from? We find again the main element at the origin of the creation of clusters: technical and technological innovations. The 20th century saw an acceleration of the pace of technological innovations along with a sophistication of these innovations. It implied simply that more and more money and better and better skilled manpower were necessary to keep the pace vis-à-vis the competition. It led at least to three consequences:

. More money to invest means an easier access to innovations for large companies and government-supported companies

. They are more than often the same companies, even in “free-market” economies. . Better skilled manpower means again large companies and a proximity to universities and research centers. Universities and research centers are again in most countries in the world fully or partially supported by governments hence under their sphere of decisions.

. Because technologies are progressing rapidly, companies do not have any more reserve of local competitiveness. They have to be competitive quick and to keep the pace. So do clusters. What took centuries to be created has now to appear in one or two decades. So clusters are no more a creation of market conditions only but more and more the result of a political decision.

What is left of the ancient clusters? Outside of a few exciting places for tourist, such as porcelain in Limoges or tailor quarters in South East Asia (Little India for example in Singapore), nothing is left from the ancient form of clusters. Some remain but definitely not anymore as a source of innovation and competitiveness for a country. So we are left with new form of clusters, which try to replicate the virtuous circles of the ancient form of clusters but by having only a few years to achieve important results. We can currently distinguish between two types of clusters: the Cold war clusters and the late 20th century clusters. What we call the Cold war clusters are the clusters which grew between the end of World War II and the beginning of the 80’s. Among those clusters, we can mention of course the Silicon Valley in the US, Sophia-Antipolis in France or the Cambridge Science Park in UK.

What are the characteristics of the Cold War clusters?

. They all have strong public research at the center of the clusters. Mostly in the form of universities, sometimes in the form of research centers only, these clusters benefits from the direct transfer of technologies from public entities. Even in cases where universities or research centers are private, they have benefited and still benefit from the support of public money in the forms of research grants or research contracts. The money which fueled most of these clusters was a direct consequence of the Cold War where governments were investing lots of money to benefit from the latest advancements in weapons technologies. These technologies were then transferred to other industries to be used in non-military products. Most of the technological innovations and most of the successful clusters in the 20th century have benefited directly or indirectly from this situation.

. These clusters are all clusters organized around technology-intensive industries. Because of the origin of the money, clusters were developed around high-tech innovations. We don’t know of any clusters developed at that time in Western Europe or the US which is not around high-tech PRODUCTS.

. They took 20 to 50 years to really express their full potentials. We have seen that the old form of clusters took decades and sometimes centuries to be established as centers of excellence. In the case of the Cold war clusters, it went faster, thanks to public support but was still not an overnight process. So it is so far a constant in the study of clusters that time is the essence of the success.

. They were all made of a mix of large and small companies. Cold war clusters have large companies as leaders. These large companies are often the first beneficiaries of public contracts. They are also the first to get access to technological innovations developed in universities and public research labs and often participated to their financing through tripartite contracts with these universities and the government. We don’t know of any successful cluster made only of small innovative companies.

. As a consequence, SMEs and start-ups are the secondary beneficiaries of technological innovations. In high-tech clusters, the most promising innovations are going first to large companies. At a second stage, large companies are using smaller ones for subcontracts which can imply the transfer of technologies.

. Because of the industries around which these clusters were developed, SMEs are almost never at the origin of disruptive innovations. Radical or disruptive innovations will again originate in public research labs or in partnerships between public labs and large companies R&D centers.

. If they want to keep the pace, these clusters have to constantly reinvent themselves. The best example is the Silicon Valley which moved from a center of excellence in the semiconductor industry to the IT industry, first hardware then software, and now trying to remain at the edge in the biotech industry and in the promising green technologies industries. This is also the only way to continue to attract public funding.

The second type of clusters, the newest one, is the clusters created at the end of 20th century and beginning of the 21st century. These clusters have more dissimilarities than similarities with the former forms of clusters. The main (and only?) similarity that we find is that new clusters are all backed up by governments. Why are clusters one of the most important tools used by governments to develop a competitive advantage? This comes mostly from the vast literature and studies made in the 80’s and 90’s on the Cold War clusters which showed that clusters appear to be a good way for a country to create competitive industries. Outside of that, there are many differences:

. First, new clusters are not always created around strong centers of public research. If this is almost a fact in the case of the old clusters, governments are creating them around strong centers of public research… when they have some in their countries. If not, they build clusters associated with universities, hoping that the mix of large companies and public centers research will lead to more innovation. Or even sometimes, if they really don’t have universities or really too young ones, they will try to import knowledge by asking foreign universities to build a campus or open an office in the newly created cluster hoping again that these foreign universities will come with some hot research or at least facilitate transfers. This situation is found mostly in countries in the Middle East and South-East Asia but also in the recent governmental initiatives to create new clusters in several European countries.

. New clusters often benefit directly from public money. Whereas for the older clusters, public money was fueling indirectly the clusters, for the newer ones, governments or public agencies would put money to build infrastructure, facilities…

. Clusters are not always created with large companies at the origin. So governments have to create incentives to attract big players, hoping to recreate the virtuous circle leading to competitiveness. As a consequence: New clusters do not attract companies, large or small, for competitiveness reasons but for financial reasons. This is certainly the most important difference between the old forms of clusters and the one. Companies (or craftsmen in the past) were interested by creating a cluster or joining an existing one because it was obvious that it would boost their innovative potentials through a better access to knowledge, to supply, to skilled manpower,…Because we believe that new clusters are mostly artificial creations of governments, most of them cannot offer the full range of business incentives. So they have to replace them by pure financial incentives, mostly in the form of various tax rebates.

. New clusters are expected to become competitive very quickly. As governments are investing a lot into the creation of new clusters, again in facilities, in tax incentives, in attracting big companies and foreign universities, they expect that the return on their investment will be very fast, and much faster than the older forms of clusters.

Posted in: Papers