As different as fintech, cleantech, and edtech are, successful clusters like these have more in common than you’d think: most benefit from geographic proximity and critical mass, which includes having a tight-knit group of stakeholders with both local and international reach. Most are complementary, featuring multisector ecosystems that share products, skills, and technologies. And most well-run clusters balance collaboration (such as joint R&D) with competition (vying for the same customers).
The power of the cluster approach is evident across different continents of the world. For instance, the Basque clusters in Spain helped the Basque Country region achieve impressive economic growth, including a 41% GDP growth, 20% increase in productivity, 7% increase in exports, and 6% decrease in unemployment rates in the first 20 years of operation. As a result, the Basque clusters outperformed the rest of Spain by one percentage point per year on GDP growth.
Similarly, the Singapore Biopolis cluster, which focuses on biopharmaceuticals, has been the source of 950 patent filings annually, the establishment of eight out of the ten largest global pharmaceutical companies, the creation of approximately 7,500 jobs, and an increased GDP contribution of pharma to 3% of Singapore's economy.
Identifying Clusters with the Most Upside: A Technical Approach
Most countries do not measure their economic data on a cluster level. Instead, their attention is too often at the sector level, treating companies within, say, the manufacturing or transportation sectors as the same. But by focusing their efforts on clusters—and then creating an environment that fosters innovation, attracts talent and investment, and builds resilience—economic planners can create a virtuous cycle of growth and success. To effectively pursue economic development through the cluster approach, countries must adopt a four-step approach to determine which clusters to prioritize.
1. Identify strongest subsectors quantitatively. The first step in building a cluster strategy is to analyze the subsectors in an economy based on quantitative data, and then rank the sectors and subsectors that hold the most promise for a particular region. (For the purposes of this exercise, sectors are areas of the economy such as manufacturing that share related activities; subsectors are subsets such as pharmaceutical manufacturing; and clusters represent a group of interconnected companies and institutions that are geographically close. Clusters can include many economic subsectors.)
The quantitative analysis can be applied at the subsector level, following three criteria and eight indicators to rank the best subsectors to anchor the economy over the next five to ten years. The three criteria consist of:
- Current economic contribution, representing the existing presence and activity of the subsectors. This includes not only the cluster’s contribution to GDP but also its success at creating skilled jobs, value-added economic activity, and exports.
- Historical investment, which measures past investments and the accumulated fixed capital. This measurement also looks at the allocation and growth of investments, including by international financiers.
- Export potential to the country’s top ten trading partners, and whether that will boost economic output.
When completed, this quantitative analysis identifies clusters that have enough presence to attain economic growth in five to ten years.
After identifying the best indicators for the criteria, the next step is to identify the relevant data, which must come from reliable sources. Data needs to be mapped to International Standard Industry Classification or a compatible format that, like ISIC, uses numerical codes to group businesses by their primary economic activity. Compatible formats include the North American Industry Classification System (NAICS), Standard International Trade Classification, and the Harmonized System.
Once the indicators have been finalized and the data sources have been identified, then the identification process can begin. In this model, there are four steps that can take the data from its original form to the end result:
First, we start by making the data easier to work with by converting it into an ISIC format.
Second, the data goes through an analysis. This involves listing and comparing measures for each category, making them easier to understand by scaling the scores from 0 to 10, based on the highest and lowest values for that category. Take, for example, the contributions from foreign direct investment (FDI). If the highest contribution in one year was 30% and the lowest was 2%, then a subsector that received 17% of FDI would receive a score of five, since it is in between the highest and lowest values.
Third, after the analysis the scores are combined using weights that the user provides. These weights are important because they reflect the importance placed on different factors. For example, if planners are focused on the next five years, then subsectors making strong economic contributions today will be given more weight. Finally, subsectors with the highest scores are classified as "priority subsectors," indicating their potential upside.
In the fourth step, the high-scoring subsectors are classified as “priority subsectors.”
2. Map priority subsectors to their relevant clusters. The second step in identifying the most promising clusters is to map the top subsectors to their relevant clusters. Fortunately, there are road maps that can help. For example, Harvard Business School’s Michael Porter and the US Cluster Mapping Project created a taxonomy of 67 economically important clusters in the United States. While this project focused on US industries, the principles and methodology can be easily adapted for use in other regions around the world.
Of the 67 clusters Porter identified, 16 were classified as “local” and the other 51 as “traded” or “tradable”—meaning that they sell locally, nationally, and even internationally. (Note that clusters can have a local and a tradable equivalent: for instance, there is a local motor vehicles cluster as well as a traded automotive cluster.) The methodology for traded and local clusters is explained on the US Cluster Mapping Project website.
Each of the 67 clusters covers multiple sectors and multiple subclusters. For instance, the Cultural and Educational Entertainment subclusters encompass many clusters including art dealers; museums; historical sites; zoos and botanical gardens; and more.
The priority subsectors serve as “anchors” for specific clusters and can be identified through the methods described in Exhibit 2. The outcome of this exercise is the identification of clusters with potential that map to these anchor subsectors.