This is the first in a three part series, in which we look at how our new industry clusters can be used to develop a sound strategy for local growth. In this part, we look at why there is a need for clustering industries together in a new way, and how we have done this. In part 2, we will look at how the clusters can be used to gain rich insight into the strengths, opportunities and threats in local economies.
Clustering industries by economic activity
With Brexit on the horizon we are entering an unprecedented and exciting period for trade and investment across the UK. Together with the devolution agenda, the end of central government funding for local councils, and increased competition between regions to attract and retain businesses and investment, it is becoming imperative for regions to articulate and evidence their “proposition” to business and those who are advising them.
However, in order to be able to do this, regional economic developers first need to ensure that they have a sound understanding of their local economy, and in particular be able to answer the following questions:
● Which industries give us a comparative advantage over other regions?
● Which sectors present us with the best opportunities to grow?
● Which industries might be under threat?
Successfully answering these questions can be used to provide a solid foundation for the development of a sound local growth strategy. However, in practice, this is by no means straightforward.
The Limitations of the Standard Industry Classifications
Perhaps the most obvious way to attempt an answer to these questions is by using local industry data. But aside from the problem of getting hold of data which is granular enough to give an accurate picture of a local economy, the big problem in using this method is the sheer number of industries. The Standard Industrial Classification system (SIC) categorises 563 different sectors at the 4-digit level, making it extremely hard to identify trends, let alone understand which industries are driving the economy.
One possible answer is contained within the SIC system itself. 4-digit classes are added up to form groups(3-digit SIC), which are then added up to make divisions (2-digit SIC), which are themselves grouped together to form sections (1-digit SIC). And so an economic developer attempting to answer the above questions could make things easier by looking at industries at the 1, 2 or 3, rather than 4-digit level.
However, the groups, divisions and sections within the SIC classification system are severely limited and don’t give a very good picture of what is actually driving an economy. The reason for this is that the higher level groupings are driven by similarity of activity, rather than by any economic linkages between them. This may be a convenient way to cluster industries together, but it doesn’t actually make very much sense when trying to understand how a local economy works. For example, if there is a large food and drink industry, the size of the total manufacturing workforce, which is what the SIC system identifies, is of far less significance than the links between manufacturing and distribution centred around food and drink.
A better solution would be to take those 4-digit SIC industries, and group them together not according to similarity of activity, but rather according to their economic links with one another. If this could be done, we would not only have a far more simplified process of identifying trends than trying to make sense of the 563 industries, but more than that we would have a far more accurate way of identifying which sectors are driving a local economy than using the SIC 1,2 or 3 classifications
The methodology behind industry clusters
If the intention is to understand the economic diversity of a region, grouping industries together according to their economic links with one another would seem to be a sensible approach. After all, in today’s connected economy, it is not isolated product and process that matter, but the connections from supply to demand. As Delgado et al. (2016) argue, “the agglomeration of related economic activity is a central feature of economic geography”¹. But how can this be done in practice?
Back in 2016, a major project was undertaken at Harvard Business School’s Institute for Strategy and Competitiveness, with the aim being to define a benchmark set of clusters. But rather than looking at similarity of activity – for example, grouping manufacturing industries together as the SIC classification system does – they instead grouped sectors together based on a number of criteria that link them together economically. These include:
● Industries which tend to co-locate in the same areas
● Industries that share a similar workforce
● Industries which tend to have supply chain connections
Emsi has applied a similar methodology to industries in the UK. Beginning with the 563 4-digit SIC
industries, we have identified those that are most economically linked according to the criteria established by the Harvard study. This has resulted in a set of 49 ‘coherent’ industry clusters.²
In addition, as we will explain in the next part in more detail, we have then divided these clusters into two distinct groups, in order to offer really rich insights into the forces driving local economic development.
¹ Classic references to agglomeration are Marshall (1920) and Krugman (1991). For the role of clusters in measuring agglomeration, see inter alia Ellison and Glaeser (1997), Porter (1998) or Delgado et al. (2014).
² You can download a list of the cluster definitions by clicking here.
If you would like to find out more about how we can help you grow your local economy, contact us at email@example.com