Economic growth

Clustering effects in London: Is it just Shoreditch or we should also be looking at Poplar?

Yannis Pierrakis - 16.05.2012

Interest in the role of external economies and spill over effects as an influence on regional and local growth has grown remarkably in recent years.

Industry clusters policy approaches are based on the logic that external economies shared by a group of co-located firms will elevate the level of competitiveness and rate of growth of the group overall. [1] Clusters, which are defined by Porter [2] as geographic concentrations of interconnected companies and institutions in a particular field, encompass an array of linked industries and other entities important to competition.  This blog aims to investigate whether there is a clustering effect of computer related companies in specific areas of London. While it is impossible to artificially constrain the sharing of tacit knowledge within postcodes, this analysis provides an indication of spatial proximity of such companies within London. [3] 

Figure 1: London based computer related companies with less than 50 employees, established between 2007 and 2011



Source: Nesta, data from FAME

Figure 1 illustrates the proportion of computer related start ups in each London postcode. This includes hardware consultancy; software consultancy and supply; data processing; data base activities; maintenance and repair of office accounting and computer machinery (only a very small proportion of companies operate in this subsector); other computer related activities. W1, EC1V and E14 is home to a relatively large proportion of computer related start ups.

The question that now arises is whether this distribution of these companies to different postcode is "fair". For example, is it fair for E14 to host around 2.14 per cent of all London computer start ups? Should we expect E14 to have more or less? In order to answer these questions there is a need to identify a measurement of fairness or what might be regarded as an expected level of new company creation.

Figure 2 presents the local distribution of young computer related companies in the form of location quotients (LQ)which indicate each postcode's share of such companies as a ratio of the proportion of young companies in postcode (measured by the number of all young companies with registered trading address in London) to the proportion of computer related start up companies in postcode. A value of over one indicates that a postcode has more than its expected share of such companies based on that region's share of the London start up population whereas a value of less than one indicates that its share is less than expected.

Figure 2: Concentration (location quotient) of London based computer related young companies by postcode



Source: Nesta, data from FAME

ECV1 (Old Street) has LQ  of 2.01 while EC2A (Shoreditch) has a LQ of 1.48 indicating that this relatively small area of London has by far more computer related young companies than expected. However, we also see that E14 (Poplar district: Poplar, Isle of Dogs, Limehouse, Canary Wharf, Blackwall, Cubitt Town) has the highest LQ (2.23) indicating a more than expected concentration of computer start ups than expected.

This finding suggests that there is concentration of computer start ups not only in the Silicon Roundabout but also in Poplar. Is this a genuine clustering effect and why it may be happening?

[1] Rosenfeld, S. 1996. Does Co-Operation Enhance Competitiveness? Assessing the Impacts of Inter-Frim Collaboration, Research Policy 25(2), p. 247-263

[2] Porter, M., 1998. Clusters and the new economics of competition,  Harvard Business Review 76(6), p. 77-90

[3] For the scope of this analysis, London based computer related  start ups are defined as follows: active companies operating in computer and related activities (UK SIC 2003:72), have registered trading address in London, have been established in the last five years (2007-2011) and employ less than 50 people.  The analysis has been conducted using the FAME database. FAME provides financial information about 2.6 million UK public and private companies. Searches can be made by company name or by criteria such as location, turnover size and number of employees.   It is important to note that the analysis includes only computer related companies (SIC code 72). This includes hardware consultancy; software consultancy and supply; data processing; data base activities; maintenance and repair of office accounting and computer machinery (only a very small proportion of companies operate in this subsector); other computer related activities. Not all computer related companies are high tech start-ups (typical Silicon Valley companies). A follow up analysis may try different methodology  in order to probe beyond SIC classifications