This article is one of a series of articles examining both the geographical concentration of industries and the industrial specialisations of local areas. This article examines regional and local industrial specialisations in Great Britain through use of employee jobs data. A number of measures of industrial specialisation have been calculated to provide a means of measuring the similarity, or otherwise, of industrial structures between areas. In addition, location quotients have been calculated. These allow for a more in-depth analysis of local specialisations. The article illustrates how the results can be used to inform analysis, and it also provides some context on the issue of specialisation. The underlying data used is from the Business Register and Employment Survey.
The North West of England has an industrial structure1 most similar to that of Great Britain overall. London has the most dissimilar.
The most similar region to London is the South East. However, to emphasise how different London is to the rest of Great Britain, in terms of industrial structure and based on the index calculated in this article, the South East is more similar to Wales, Scotland and all other regions of England than it is to London.
At the local authority level, the areas with industrial structure most similar to Great Britain overall are Leeds, Birmingham and Bristol. The most dissimilar areas are City of London, Isles of Scilly, North Warwickshire and Tower Hamlets.
An excel tool (2.92 Mb Excel sheet) has been included in the article which allows users to investigate for each local authority the local authorities with the most similar and dissimilar industrial structures.
The article also includes an excel tool (1.96 Mb Excel sheet) that allows users to examine industry location quotients for each local authority. This is a standard way of investigating an area’s industrial specialisations. As an example, for Cheshire West and Chester the data illustrates the local authority’s diverse specialism in chemical manufacture and the monetary intermediation sector.
Possessing industrial specialisations can be positive for a local area. A local concentration of economic activity in an industry brings incomes and employment to an area as well as generally having positive productivity effects for the economy overall. But too much reliance on a small number of industries can be problematic if there is a negative economic shock to one of these industries.
An accompanying article that provides detail about the spatial concentration of industries is also available to view.
Analyses of industrial structures in this article are based on a comparison across different geographical areas of employee jobs data split by industry.
This section examines how similar or dissimilar an area’s industrial structure (as measured via employee jobs data) is relative to the structure of the economy of Great Britain overall. This has been done through calculation of a number of indices of specialisation1. The results shown in the tables below utilise the Krugman index. Additional results based on the Gini Coefficient, Theil index and the Herfindahl index are available in the accompanying data tables2.
The Krugman index compares the industrial structures, in this case measured by the split of employee jobs across industries, between two areas. The more similar the industrial structure the lower the index and the more dissimilar the industrial structure the higher the index.
Table 1 shows the results when the industrial structure of Scotland, Wales and the English regions are compared in turn to the overall industrial structure of Great Britain.
|Yorkshire and The Humber||0.21|
The higher the Krugman Index the more dissimilar is the industrial structure to that of Great Britain overall.
London is shown as having an industrial structure most dissimilar to that of Great Britain overall. This is not a surprising result as London is well known for specialising in industries such as finance, business services and media that are less common elsewhere in Great Britain.
Aside from London, it is Wales and the North East of England that are shown as having industrial structures most dissimilar to GB overall. Meanwhile, the North West region is shown as having an industrial structure most similar to that of Great Britain overall.
Table 2 shows a similar set of results but this time based on an analysis of industrial structure across Local Authorities. The table shows the twenty local authorities that have the highest and lowest results on the Krugman index. In other words, local authorities whose industrial structure is most dissimilar to Great Britain overall followed by those most similar.
|Most Dissimilar (High Relative Specialisation)||Most Similar (Low Relative Specialisation)|
|City of London||London||Leeds||YH|
|Isles of Scilly||SW||Birmingham||WM|
|North Warwickshire||WM||Bristol, City of||SW|
|West Somerset||SW||Cheshire West & Chester||NW|
|Kensington and Chelsea||London||North Somerset||SW|
City of London is shown as having an industrial structure most dissimilar to that of Great Britain overall. A further four London boroughs including Tower Hamlets and Westminster are also within the list of twenty local authorities with the most specialised industrial structures (relative to the Great Britain average). Other local authorities in the list are spread across the country with local authorities in a further six English regions included in the list together with two local authorities in the Highlands and Islands of Scotland.
The local authority with an industrial structure most similar to that of Great Britain is seen to be Leeds, followed by Birmingham and Bristol. All the local authorities in this list are within England.
The majority of the regions and countries of Great Britain have local authorities amongst both those most similar to Great Britain overall and those with the largest relative specialisation. In other words, there is not one region that dominates either list. This suggests that most regions and countries have quite a mix of different industrial structures amongst their constituent local authorities, including some that have industrial structure similar to Great Britain overall and some that have particular specialisms that make their industrial structures more distinct.
The results and rankings for all Great Britain local authorities can be found in the accompanying data tables. In addition to showing the results for the Krugman index, they also show results calculated via alternative relative measures of specialisation, namely the Gini and Theil indices. In each case these calculate results relative to the industrial structure of Great Britain overall.
An alternative way to investigate specialisation is via an absolute measure. Such an index compares the industrial structure in each region to a uniform distribution. In other words, it compares to a case where employee jobs are equally split across all industries. The Herfindahl index is an example of an absolute measure of specialisation and its results are also included in the accompanying data table.
The results are sensitive to the level of industrial structure used. A more disaggregated split of industries will give a better result and these results are based on an analysis using a 272 industry split. However, it should be noted using a different choice of industrial split would likely lead to slightly different results.
For each of the indices the only data being used in these comparisons are employee jobs data by industry from the Business Register and Employment Survey. The indices are comparing how a regions employee jobs split by industry compares to either Great Britain overall or to a uniform distribution. It should be noted, however, that the actual occupations within industries in different areas may be quite different, for example an area may have relatively high or low skill occupations in a particular industry when compared to Great Britain. This analysis does not account for this or for other structural economic issues such as levels of unemployment or economic inactivity in an area.
The previous section examined the industrial structures of regions and local authorities relative to the overall economy of Great Britain. However, it is also possible to investigate how different local areas compare to each other. This can be useful in providing an indication of which other areas have a similar industrial structure and which do not. Such knowledge may be useful, for example, in benchmarking exercises. As in the previous section, the Krugman index has been used to make these comparisons.
|North East||North West||Yorkshire & The Humber||East Midlands||West Midlands||East of England||London||South East||South West||Wales||Scotland|
|Yorkshire & The Humber||0.33||0.24||0.26||0.25||0.31||0.55||0.34||0.28||0.29||0.35|
|East of England||0.40||0.29||0.31||0.28||0.27||0.52||0.25||0.29||0.38||0.41|
Table 3 shows that the two regions in Great Britain with the most similar industrial structures are the North West and Yorkshire and The Humber. Other areas that have similar structures include the South East and East of England; Yorkshire and The Humber and West Midlands; and also the South West and North West.
The most striking finding in Table 3, however, is that London is the most dissimilar region to every other region and country in the list. The values of the Krugman Index for comparisons to London range from 0.43 with the South East (most similar to London) to 0.65 to Wales (least similar to London).
By contrast, all other regions and countries are relatively similar to each other. Indeed, the two most dissimilar regions outside of London are the North East and the South East of England with a Krugman index of 0.41. However, this still makes these two regions more similar to each other in terms of industrial structure according to this index than London is to any other region/country of Great Britain.
A similar analysis has been carried out for Local Authorities. This produces a 380 x 380 matrix, in which each local authority is compared to each other local authority and a value of the Krugman index calculated. To bring some order to these 144,400 results, an excel tool has been created which allows users to select a local authority and then to see a list of the 15 most similar and 8 least similar local authorities.
Figure 1 shows a reduced form of these results, showing the top 8 most similar local authorities to Bradford and the 4 least similar. According to the calculations, the most similar industrial structure to Bradford is found in nearby Kirklees, with other local authorities in either Yorkshire and The Humber or the North West regions also shown in the chart as having a similar industrial structure (based on the split of employee jobs across industries).
The excel tool can be opened by clicking on the spreadsheet link above. This will allow the results to be produced for all other local authorities in Great Britain.
The results should be treated as being indicative. The base data being used is employee jobs split by industry for each local authority. So a similar local authority in the chart above and in the preceding analysis in this article is one in which the two local authorities have a fairly similar break down of employee jobs across the 272 industrial sectors used in the analysis.
However, the analysis does not take any account of the occupation type or quality of job on offer within each industrial sector. Additionally, this data analysis is not considering other issues related to the economic structure of an area such as unemployment or inactivity rates. Even accounting for these caveats, however, the data can provide a useful tool for discovering which local authorities have relatively similar industrial structures and which do not.
So far the article has focused on showing those areas whose industrial structure most differs from that of Great Britain and in providing an excel tool that gives an indication as to which other areas may possess a similar industrial structure. How does this link to specialisation? Broadly, areas with particular specialisms are likely to be seen as more dissimilar to either Great Britain overall or to most other local areas.
Therefore, the data in the preceeding sections gave an initial indication of those regions and areas most likely to possess industrial specialisations. The excel tool then allows an investigation of what other areas may broadly share these specialisations.
However, the analysis to date is largely indicative in that it is attempting to identify in a single number the extent that any particular area is likely to be similar or different to Great Britain overall or to other areas. Such a process can only be an approximate one. To more fully understand the industrial structure of an area it is necessary to examine more closely the disaggregated data. One common way to do this is to calculate location quotients. This is the approach taken in this section.
Location Quotients compare the industry share of employee jobs in a local area compared to in Great Britain overall. For example, if 5% of employee jobs in Great Britain occur in Industry Z, then any local authority which also has 5% of its employee jobs in that same industry will have a location quotient for Industry Z of 1.0. If the local authority has more than 5% of its employee jobs in Industry Z then its location quotient for that industry will be >1.0. Therefore, the higher the location quotient the more a local area has a specialisation in that industry relative to Great Britain overall.
Table 4 shows the Location Quotients calculated for each of the Industrial Sections for the Local Authority of Cheshire West & Chester. By opening the spreadsheet below the table, it is possible to examine similar data for all other local authorities in Great Britain using the simple menu tool.
|K : Financial and insurance activities||1.9|
|C : Manufacturing||1.2|
|I : Accommodation and food service activities||1.2|
|G : Wholesale and retail trade; repair of motor vehicles and motorcycles||1.1|
|S : Other service activities||1.1|
|P : Education||1.1|
|M : Professional, scientific and technical activities||1.0|
|E : Water supply; sewerage, waste management and remediation activities||1.0|
|F : Construction||1.0|
|L : Real estate activities||0.9|
|N : Administrative and support service activities||0.9|
|O : Public administration and defence; compulsory social security||0.8|
|Q : Human health and social work activities||0.8|
|H : Transportation and storage||0.7|
|R : Arts, entertainment and recreation||0.7|
|J : Information and communication||0.7|
|B : Mining and quarrying||0.5|
|D : Electricity, gas, steam and air conditioning supply||0.2|
|A : Agriculture, forestry and fishing||0.2|
The data for Cheshire West and Chester shows that its main specialisation relative to Great Britain overall occurs in section K: Financial and insurance activities. By contrast, it has a relatively low share of its employee jobs, relative to the shares in Great Britain overall, in the industry sections D: Electricity, gas, steam and air conditioning supply or A: Agriculture, forestry and fishing.
A more detailed examination of local specialisations can be seen from observing the same location quotient data at a higher industrial disaggregation. In the workbook, data is also presented for both a 2-digit SIC industrial breakdown (84 industry split) and a 3-digit SIC industrial breakdown which splits employee jobs across 257 industries.
Table 5 lists those 3-digit SIC industries for which Cheshire West & Chester has a location quotient of above 1.5. It shows that the three industries for which it has the largest relative specialisations are all manufacturing industries, with a particularly strong specialisation in chemicals industries. Both monetary intermediation and activities auxiliary to financial services are also in this list which confirms the finding in table 4 that the financial sector is also an area of specialism within this local authority.
|205 : Manufacture of other chemical products||8.6|
|201 : Manufacture of basic chemicals, fertilisers/nitrogen compounds, plastics/synthetic rubber in primary forms||7.0|
|109 : Manufacture of prepared animal feeds||3.9|
|641 : Monetary intermediation||3.0|
|742 : Photographic activities||2.9|
|323 : Manufacture of sports goods||2.3|
|105 : Manufacture of dairy products||2.3|
|661 : Activities auxiliary to financial services, except insurance and pension funding||2.2|
|771 : Renting and leasing of motor vehicles||2.1|
|910 : Libraries, archives, museums and other cultural activities||2.0|
|421 : Construction of roads and railways||2.0|
|453 : Sale of motor vehicle parts and accessories||1.9|
|383 : Materials recovery||1.8|
|553 : Camping grounds, recreational vehicle parks and trailer parks||1.8|
|463 : Wholesale of food, beverages and tobacco||1.6|
|451 : Sale of motor vehicles||1.5|
|477 : Retail sale of other goods in specialised stores||1.5|
|476 : Retail sale of cultural and recreation goods in specialised stores||1.5|
|551 : Hotels and similar accommodation||1.5|
|631 : Data processing, hosting and related activities; web portals||1.5|
Again by opening the Excel workbook, the data can be examined for all other local authorities. This data can help provide a good understanding of the industrial sectors in which a local authority specialises.
Once a sector has been identified as a specialisation it may be that there is interest in knowing which other local authorities in Great Britain have a similar specialisation. To this end a companion article, looking at the issue of industrial concentration, provides a useful data tool (1.93 Mb Excel sheet) for investigating in which areas of the country location quotients are high for each industry. Maps of location quotients for many industries are also included with the companion article which is entitled 'The Spatial Distribution of Industries'.
Finally, in this section, a couple of cautions for when interpreting the location quotient data presented above. Firstly, it should be noted that the location quotients only show a relative picture compared to Great Britain so that one cannot infer from a high location quotient alone that there are a large absolute number of people working in that industry within the local authority.
If an industry is relatively small in employee jobs terms in Great Britain overall, it is possible for an area to have a high location quotient for that industry but only a relatively small number of actual employees. Meanwhile, in an alternative sector it may have a much lower location quotient despite having a much larger number of people employed in that sector – the reason being that this alternative sector is a large employer across Great Britain overall.
So the key point is that the size of the location quotient is a relative calculation that is partly based on the overall size of the industry in Great Britain as well as the local level of employee jobs in that industry.
Secondly, for some industries it has not been possible to calculate location quotients for all areas because some industry data is disclosive in some local authorities. This can happen where employee jobs are concentrated in only a small number of firms. In many cases, the number of employee jobs in such a case are likely to be low. However, where an industry in an area is dominated by one or a small number of very large plants or firms, then there may be a specialisation in that area which it was not possible to show in the results due to this issue of data disclosure.
Local knowledge may help isolate examples where this has been the case. Also, if the data is unavailable at the SIC 3 level, it may be available at a slightly less disaggregated breakdown. For example. in the case of Cheshire West & Chester, at the SIC 2 level the Manufacture of motor vehicles, trailers and semi-trailers (SIC 29) has a location quotient of 4.24 suggesting a significant specialisation in motor vehicle manufacture in the area. However, for disclosure reasons, the data for SIC 291: Manufacture of motor vehicles was not available and therefore wasn’t able to be included in Table 5 above.
Specialisation occurs in conjunction with trade. The ability to trade goods and services whether locally, nationally or internationally means that individuals and firms are able to specialise in the production of a particular good or service. Furthermore, for many goods and services it is possible to transport the end-product significant distances meaning location decisions for plants and offices are not dependent on customer location.
Instead, firms choose to locate in particular destinations for a myriad of reasons and these location decisions often result in a spatial concentration of firms from particular sectors. The flipside of this is that the area that is home to this concentration of firms in a particular sector will likely have a specialisation in this particular industry relative to the rest of the country.
At a local level, industrial specialisation can be seen as potentially having both positive and negative effects. As discussed in the accompanying articles on industrial concentration, there are wider benefits to the economy overall of spatial concentrations of industrial output through potential productivity gains. However, for any particular local area, the impact on standards of living will be determined by which industries are specialised locally. In general, an area with specialisms in high productivity sectors is likely to have higher incomes amongst its residents than an area with specialisms in lower productivity sectors.
The negative effect of specialisation can occur if an area is highly dependent on a particular industry and that industry is in gradual or sudden decline. As an example, a local area can be highly impacted by the closure of a major steel works or car plant, via its effect on the local area in terms of reduced employment and output and any knock on effects on other local businesses.
As such, local policy makers are often simultaneously interested in policies around clusters and encouraging specialisms to their areas to attempt to gain the benefits that can accrue from such a specialism whilst also being interested in their areas ‘resilience’, namely how well the area would cope with negative shocks to the economy and particularly to any existing local specialism. As such, an ideal situation is often considered to be one where an area has a number of different specialisms, rather than being heavily reliant on a single industry.
In the absence of production and trade data at sub-national level, employee jobs data can be used to measure the level of industry specialisation at local areas. The analysis in this report uses the number of employees by industry for regions and local authorities derived from the Business Register and Employment Survey (BRES) 2011. This means that self-employed jobs, HM Forces and Government supported trainees are not included in the data.
The BRES survey is ONS’ primary source for jobs estimates at a detailed geographical and industrial level. It contains information on the number of employees by industry down to five-digit industrial classification as defined in the UK Standard Industrial Classification (SIC 2007) and by geographic area down to local authority.
Estimates are subject to sampling error, which increase as geographic areas become smaller and industry classification become more detailed. In this article, analysis is limited to local authority level data and to the 272 industry split, i.e., the three-digit code industry level from the UK Standard Industrial Classification 2007.
The term ‘employee job’ is used throughout this article. It is used to help stress that all the data in this article is focused on the workplace of an employee. This contrasts to some data sources that focus on the residential status of an employee. However, it should be noted that strictly speaking the BRES data being used in this article is not a count of the number of jobs filled by employees in an area, but rather a count of the number of employees who work in an area.
The Standard Industrial Classification (SIC) provides a framework for the collection, tabulation, presentation and analysis of data, and its use promotes uniformity. The UK SIC is a hierarchical five digit system. UK SIC (2007) is divided into 21 sections, each denoted by a single letter from A to U. The letters of the sections can be uniquely defined by the next breakdown, the divisions (denoted by two digits).
The divisions are then broken down into groups (three digits), then into classes (four digits) and, in several cases, again into subclasses (five digits). There are 21 sections, 88 divisions, 272 groups, 615 classes and 191 subclasses. In this article, results are shown for the industry sections, divisions and groups only. The full SIC breakdown can be found in the UK SIC structure and explanatory notes document.
The rest of this section details the different indices used in this article. It explains each of the measures of specialisation used together with an explanation of location quotients.
The Krugman index is regarded as the standard index among relative specialisation measures (Palan, 2010) and is therefore used in the analysis in this article in preference to the gini coefficient or theil indices. The Krugman index is sometimes referred to as the Krugman Similarity Index or the Krugman Dissimilarity Index.
The Krugman index compares the industrial structures of two geographical areas. It runs from zero, if area A has the same employee jobs split across industries as the reference area, to two if they have employee jobs only in entirely different industries to each other. Letting Ei,r be the share of industry i in area r, then the Krugman Index is calculated as follows:
The reference area can vary. In the section of this article “Comparison to the Industrial Structure of Great Britain “, the reference area is set as being Great Britain such that the industrial structure of local authorities and regions were each compared to the industrial structure of Great Britain. In the section of this article “Comparisons between local areas”, the reference area was set in turn for each region and local authority to allow a full comparison between all local authorites (and also between all regions) to be carried out.
The Gini Coefficient is another relative indicator that determines the degree of specialisation of an area in comparison to a reference structure. The reference structure used for the Gini coefficient in this article is the distribution of employee jobs across the industries in Great Britain. The Gini coefficient, widely used in the analysis of income inequality, has more recently been adopted by economic geographers in the analysis of the geographic distribution of the economic activity. The expression for the Gini coefficient for specialisation of area r is given by
where Si(j) is the share of local employee jobs in industry i(j) and Xi(j) is the share of national employee jobs in industry i (j). The locational Gini takes values between 0 and 1. The coefficient takes the value zero if the area’s employee jobs are located across industries in the same proportion as total employee jobs for the reference area. The coefficient takes values greater than zero if the distribution of the area’s employee jobs is more skewed than that of total employee jobs.
The Theil index is from the class of entropy indices and is an alternative relative index of specialisation.
Where Ei,r is the share of industry i in area r,
The Herfindahl-Hirschman index (HHI) is a commonly used absolute measure of specialisation. The Herfindahl-Hirschman index for specialisation compares the distribution of employee jobs in one geographical area with a uniform distribution in which employee jobs are equally spread across all the industries used in the analysis. The HHI for specialisation is given the following expression:
where Ei,r/Er is share of the industry i’s employee jobs in area r. The value of the HHI increases with the degree of specialisation reaching its upper bound of 1 when all the employee jobs in the area are concentrated in one industry. HHI takes the lowest value 1/K, where K is the number of industries, when the area’s employee jobs are evenly distributed across all the industries.
As an absolute measure, this indicator displays a weighting towards large industries. It would work best for equally sized industries, but industries are not equally sized and as such, industries with larger shares of employee jobs, have a larger influence on the indices value.
The location quotients are a simple and very common measure used to assess both geographical concentration of industries and industrial specialisation of regions. Used in the analysis of industrial specialisation, the location quotients compare for each industry, the industry’s share of local area employee jobs with its share of total employee jobs. The location quotients expression for industrial specialisation is given by:
where Ei,r are employee jobs in industry i in region r, Er is the total number of employee jobs in region r, Ei is total employee jobs in industry i and E is the national total of employee jobs. A location quotient of 1 indicates that the local share of employee jobs in an industry is equal to the local share of total employee jobs.
If the local share of employee jobs in industry i (Ei,r/Er) is greater than the industries share of total employee jobs (Ej/E) across the economy, than the location quotient takes values greater than 1, indicating a relative concentration of industry i in region r.
A location quotient higher than 1.0 will occur if an industry makes up a higher share of employee jobs in a specific area than that industry does nationally. For example, the finance and insurance activities sector is responsible for 32.0% of employee jobs in Tower Hamlets but only 3.9% of employee jobs in Great Britain. As a result, the location quotient for financial and insurance activities in Tower Hamlets is (32.0/3.9 = 8.1).
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: firstname.lastname@example.org