The level of public sector employment in the UK economy has increasingly become a policy issue over recent years. To help inform the debate, ONS produces quarterly public sector employment data that measure the total level of public sector employment at the national and regional level. At present, however, there have been relatively few data releases or articles examining the subregional distribution of public sector employment. Furthermore, this distribution is of interest to policy makers both nationally and locally. This article seeks to fill this gap by providing an overview of some of the data available from ONS that can be used to examine the subregional distribution of public sector employment and also the subregional distribution of private sector employment.
This article provides a number of measures to illustrate the subregional distribution of public and private sector employment in the UK. It utilises data from the Business Register Employment Survey (BRES) and from the Annual Population Survey (APS).
The article includes data on employee densities and employment rates. Employee densities provide a means of examining the location of public and private sector employee jobs. They do so by comparing, for each subregion, the number of public and private sector employees whose work is located in the subregion to the size of the subregions 16 to 64-year-old population.
Employment rates, by contrast, don’t look at the location of employment. Instead, they provide details of the actual employment situation of residents living in a subregion. They do so by measuring the share of 16 to 64-year-old residents in a subregion who report that they work in the public or private sector.
One reason for providing both employee density and employment rate data in this article is that it helps illustrate the fact that high employee densities and high employment rates are not always linked. This is because the availability of jobs in a subregion is not the only factor that impacts upon a subregion’s employment rate. Commuting patterns and the skills of residents are among the other factors that impact upon the employment rate of a subregion’s residents in both the public and private sectors.
The article provides data for all UK local authorities and NUTS 2 subregions as well as for English Local Enterprise Partnerships (LEPs). In each case, tests of statistical significance have been carried out to determine whether a subregion’s public sector employment rate and/or private sector employment rate are statistically above or below the UK average.
A statistical test of clustering has also been carried out to provide information on the spatial spread of public and private sector employment rates. The results show that there is more clustering of private sector employment across the UK than of public sector employment.
An explanation of this is that many jobs in the public sector are in health, education or public administration and these jobs are likely to be spread relatively evenly across the country in order to serve local populations. By contrast, private sector firms in some industries are not required to locate adjacent to their customers and this freedom to choose where they locate often results in clustering of private sector firms in certain areas of the country.
This is seen in England where high private sector employment rates occur most often in subregions in the south east of England and low private sector employment rates are found more commonly in the north or midlands. In comparison, local authorities with high (or low) public sector employment rates are spread more evenly across England.
In Scotland, private sector employment rates across local authorities vary widely with some subregions well above the UK average and others well below. In Wales, private sector employment rates are below the UK average for almost all local authorities. Local authorities in both Scotland and Wales typically have public sector employment rates above the UK average.
The data used in this article are for calendar year 2010. The data can be seen in more detail in the accompanying data tables and a series of Google maps showing the 2010 data are also available to view.
In order to achieve a full understanding of the extent and geographical spread of subregional public sector employment in the UK it is necessary to consider a number of different statistics. This is because it is necessary to consider the differences between measures of employees and employment and between workplace based measures and residence based measures. Furthermore, there are a number of different comparators against which it is valid to analyse this data. For example, to allow comparisons across areas, the level of public sector employment can be expressed as a share of total employment or alternatively as a share of (aged 16 to 64 years) population.
This article shows how by combining a number of different statistics covering both the public sector and private sector it is possible to build a more rounded picture of the levels of public sector employment within subregional economies than is the case from relying on one statistic alone.
ONS’s preferred measure of regional public sector employment data are the PSE statistics. The PSE statistics are a workplace-based measure compiled from surveys and administrative sources to provide a definitive count of the level and location of public sector employment. However, the PSE statistics are not compiled subregionally.
Therefore, for subregional analysis, this article utilises data from the Business Register Employment Survey (BRES) and from the Annual Population Survey (APS). BRES is a workplace-based measure of employees1 and the APS is used in this article as a residence2-based measure of employment. All data used in this article is for calendar year 2010.
A workplace based measure of employment focuses on the location of the employment rather than the residential location of the person employed.
For a workplace analysis at the subregional level, the recommended data source is the annual Business Register Employment Survey (BRES). This provides a measure of the number of public sector employees and private sector employees working in an area. BRES is based on survey returns from businesses and as such is regarded as providing an accurate split of public sector compared to private sector employees.
An employee is defined as anyone aged 16 years or over that is paid directly from the payroll, in return for carrying out a full-time or part-time job or being on a training scheme. Individuals are classified to the public sector in BRES dependent on the legal status of the organisation that they work for, as classified on the Inter Departmental Business Register, based on UK National Accounts concepts and definitions.
It should be noted that the BRES data used in this analysis is for employee jobs only. Self-employment and employment in HM Armed Forces are not covered.
A residence-based measure of employment focuses on the residential location of the person employed rather than the location of their employment.
ONS produces the Annual Population Survey (APS) which is a residence based survey that asks individuals about their employment situation and includes a question that asks whether they are employed in the public or private sectors. The APS provides data down to local authority level.
It should be noted that estimates of the number of people working in the public sector from the APS are generally higher than ONS's official estimates of public sector employment (PSE). For Jan–Dec 2010, the PSE statistics show 21.5 per cent of UK employment in the public sector and 78.5 per cent in the private sector. The corresponding figure calculated from the APS is 25.4 per cent for the public sector and 74.6 per cent for the private sector.
This over-estimation is because the APS relies on individual respondents or their proxies classifying themselves and this often leads to individuals who work within public sector premises, while being employed by private sector organisations, misclassifying themselves as public sector workers in responding to this survey. As a result the APS over-estimates public sector employment and under-estimates private sector employment.
In terms of the results shown in this article, the 2010 UK public sector employment rate for 16 to 64-year-olds calculated via the APS is 17.7 per cent and the private sector employment rate 51.9 per cent (see Table 2 in results section). However, if respondents to the APS had reported themselves as working in the public or private sector in the same proportions as the PSE statistics, then the equivalent employment rates would be 15.0 per cent and 54.7 per cent. This illustrates that the misclassification of private and public sector employment in the APS is far from negligible.
However, a key reason for still using this data is that it can be assumed that the misclassification within the APS is evenly distributed across areas. As such, the APS data provides a very useful source for comparisons of relative public sector (and private sector) employment across different subregions. It is also the only data source that can provide a residence-based measure of employment, rather than a workplace-based measure.
Table 1 gives a summary of the two data sources in relation to subregional public and private sector employment data. It highlights that there are some disadvantages to note related to each of the data sources, but that between them they provide a range of possibilities for analysis.
|Business Register Employment Survey (BRES)||Annual Population Survey (APS)|
|Type of employment data||Workplace-based measure of employee jobs1 in a location.||Residence-based2 measure of the employment status of residents in a location.|
|Coverage||Employees only.||Employees, self employed, unpaid family workers and government supported trainees.|
|Accuracy of public/private sector split||Good||Over-reporting of public-sector employment and under-reporting of private sector employment.|
|Measures used in this article||Public sector employees as a proportion of total employees; Public sector employee density; Private sector employee density.||Public sector employment rate; Private sector employment rate.|
The data used in this article are for calendar year 2010 and a time-series of these data has not been compiled. BRES data are only published annually and the first year of publication was 2008 so there are limits to the amount of time series information that could be derived from this source. The APS, by contrast, is published quarterly and ONS will examine providing further subregional data on the APS measures if there is user demand for this.
The following links provide further information on ONS public sector employment data and the data sources used in this article,
BRES also produces a measure of employment, which is the sum of employee jobs plus the number of working owners who receive drawings or a share of the profits but are not paid via PAYE. However, this article concentrates on using the more widely used employees measure.
Five different statistics are provided in this article. The first three examine the level of public sector employment in each subregion. In some cases the three measures will all correspond such that they all show a similarly high or low level of public sector employment for a particular subregion. However, the effects of commuting can mean this is not always the case.
Furthermore, the first measure below can be heavily influenced by the level of private sector employees working in an area. Therefore to investigate the effect of the level of private sector employment on the overall employment situation in each subregion, two measures of private sector employment are also provided.
Further details of each measure being used are as follows:
Public sector employees / (public sector employees + private sector employees).
This is calculated using workplace data from the Business Register Employment Survey (BRES). It is a commonly used measure and can identify areas which are the location for a high or low share of public sector employee jobs relative to total (public sector + private sector) employee jobs.
A disadvantage of this measure is that the results are dependent on the level of private sector employee jobs in an area (which is part of the denominator). In other words, a high ratio for this measure may be due to a ‘high’ level of public sector employee jobs in an area. But equally, it could be due to a ‘low’ level of private sector employee jobs in an area, or to a mixture of the two. From this statistic alone, we cannot tell which of these alternatives is the case.
Public sector employees / population aged 16–64.
For example, a public sector employee density of 20 per cent for a subregion would signify there were 20 public sector employees working in the subregion for every 100 residents aged 16 to 64 who are resident in the subregion.
Note that the employees whose jobs are in the subregion will not necessarily be residents of the subregion. Many employees may in-commute from other subregions. Because of this a high employee density in a subregion does not directly imply a high level of employment among the subregion’s residents. If a subregion attracts a high level of in-commuting it may have a high employee density but still have a low employment rate amongst its residents.
Residents aged 16–64 employed in the public sector / population aged 16–64.
For example, a 20 per cent public sector employment rate implies that 20 residents out of each 100 residents aged 16 to 64 in an area are employed in the public sector.This measure is not concerned with indicating how many jobs are located in the subregion. It is only telling us how many residents are in employment. This employment may or may not occur within the subregion. A subregion with a high level of out-commuting may have a high employment rate amongst its residents despite having a low employee density within the subregion.
Private sector employees / population aged 16–64.
This is the same as measure 2 except that it is measuring private sector employee density rather than public sector employee density.
Residents aged 16–64 employed in the private sector / population aged 16–64.
This is the same as measure 3 except that it is measuring a private sector employment rate rather than a public sector employment rate.
The article provides data and analysis first for UK NUTS 2 subregions2 and then for local authorities in Great Britain (data for Local Enterprise Partnerships in England are given in the accompanying data tables but not discussed in the text). The NUTS 2 data benefit from larger sample sizes than other subregional geographies and therefore will be the subregional data with the most reliable results. As such it is useful for obtaining some headline results for the geographical spread across the UK.
Local authority estimates, by contrast, are formed from lower sample sizes and therefore the confidence intervals surrounding the data are larger. This means more care must be made in interpreting the results at this level. However, from an administrative perspective local authority data is a more useful geographical level than NUTS 2 data so they are included here, albeit subject to some caveats in terms of data quality.
For each subregion, the sum of the public sector employment rate and the private sector employment rate calculated in this article is slightly lower than the subregion’s overall employment rate as published in the monthly ONS release: Regional Labour Market Statistics. This is because there are a small number of people (0.5 per cent of 16 to 64 year olds across the UK) that the Annual Population Survey reports as being in employment but for whom it is does not report their public or private sector status.
At the regional and national level ONS recommends the use of PSE statistics to measure public sector employment. However, PSE statistics are not available at the subregional level.
At the subregional level, this article has introduced five different measures (a table of the results can be found in the ‘Results and analysis – NUTS 2 subregions’ section as well as in the accompanying data tables). It needs to be stressed that the five measures are not easily comparable for a single subregion, in that they measure different concepts and use different data sources with different definitions. The best method for analysis is therefore to analyse each measure individually with reference to the UK or GB average. By doing so one can compare the results found from analysing the five measures individually to build an overall picture.
To illustrate this, short analyses on a number of the subregions are provided as examples.
The first example shown in Figure 1 is for the NUTS 2 subregion of Merseyside. The figure shows the percentage point difference from the UK average for each of the measures.
The remaining columns fill in the picture. They show that the public sector employment rate for Merseyside residents is above average (2 percentage points above the UK average) while the private sector employment rate is sharply below average (8 percentage points below the UK average).
The net effect of this is that the total employment rate on Merseyside is around 6 percentage points below the UK average. Overall, therefore, the picture for Merseyside is one of a relative shortage of private sector employment, partially compensated for by a slightly above average public sector employment rate, but mainly reflected in a low total employment rate among its 16 to 64-year-old residents.
Meanwhile, the fact that the employee density and the employment rate data show similar differences from the UK average for both the public sector and private sector suggests that net-commuting does not have a major impact on the residential employment rates in this particular subregion.
The second example is the NUTS 2 subregion of Hampshire and Isle of Wight. In this subregion, the share of public sector employees working in the subregion as a proportion of total employees working in the subregion is below the UK average. However, this is not due to there being a below average public sector employment rate in the subregion, for as Figure 2 shows the employment rate in the public sector within this subregion is above the UK average.
However, while its public sector employment rate is only slightly lower than that in Merseyside discussed above, its private sector employment rate is significantly higher. Overall therefore, Hampshire and Isle of Wight is an example of a subregion that has above average employment rates in both the public and private sectors, showing that they are not always mutually exclusive. The NUTS 2 subregion of West Midlands is an example of where the opposite is true and both public and private sector employment rates are below the UK average.
The final example included in this article is the NUTS 2 subregion of Outer London, as it is a good example of how commuting can impact the data. Examining Figure 3 shows that there is a large difference between the employee density results and the employment rate results. Employee density for the private sector is 11 percentage points below the UK average, showing there are a relatively low number of private sector employees working in the Outer London subregion relative to its population. However, despite this the private sector employment rate in Outer London is around the UK average. The reason, of course, is the large job market adjacent to the subregion in Inner London which allows Outer London residents to gain employment via commuting.
Similar patterns are found in other subregions that exhibit out-commuting. In such subregions employee density is often below average, while employment rates are often above average.
This section examines what the data show about the spatial location of public and private sector employment when compared across subregions. It uses NUTS 2 subregional data. A further section repeats the process using local authority data. Results for Local Enterprise Partnerships (LEPs) in England are given in the accompanying data tables but not discussed in the text.
Table 2 presents the results for each of the five measures used in this analysis for each of the NUTS 2 subregions in the UK. (Details on each of the five measures used and their interpretation can be found in the sections ‘Data Sources’ and ‘Measures of public and private sector employment’).
|NUTS 2 Subregion||Public Sector Employees as a Share of Total Employees1||Public Sector Employee Density2||Public Sector Employment Rate3||Private Sector Employee Density4||Private Sector Employment Rate5|
|Tees Valley and Durham||27.9||15.3||18.5||39.6||46.5|
|Northumberland and Tyne and Wear||29.2||18.0||20.3||43.6||45.6|
|East Riding and North Lincolnshire||25.4||15.4||17.0||45.0||51.3|
|Derbyshire and Nottinghamshire||25.4||15.7||18.4||46.2||50.5|
|Leicestershire, Rutland and Northamptonshire||18.7||12.7||15.7||55.0||55.6|
|Herefordshire, Worcestershire and Warwickshire||19.2||12.9||16.9||54.5||56.3|
|Shropshire and Staffordshire||22.5||14.0||16.2||48.2||54.8|
|Bedfordshire and Hertfordshire||20.1||13.2||17.0||52.4||57.2|
|Berkshire, Buckinghamshire and Oxfordshire||16.4||12.3||17.4||62.7||57.8|
|Surrey, East and West Sussex||18.0||12.0||15.5||54.7||59.8|
|Hampshire and Isle of Wight||20.7||13.8||18.8||52.9||54.5|
|Gloucestershire, Wiltshire and North Somerset||22.7||16.1||18.2||54.9||57.2|
|Dorset and Somerset||18.0||12.4||18.1||56.1||54.5|
|Cornwall and Isles of Scilly||23.6||13.7||15.5||44.4||53.0|
|West Wales and The Valleys||31.6||17.3||20.5||37.5||43.9|
|North Eastern Scotland||19.8||16.4||18.5||66.6||60.4|
|South Western Scotland6||27.6||17.8||20.3||46.7||46.6|
|Highlands and Islands 6||34.3||23.6||24.8||45.2||52.4|
An interesting example shown in Table 2 is Inner London where public sector employee density in Inner London is one of the highest in the UK. However, its residential public sector employment rate is the lowest in the UK. This extreme example illustrates that the existence of a high level of public sector employee job opportunities locally does not always equate to a high public sector employment rate for residents (this can also be true for private sector employee density and employment). Commuting patterns and the skills of local residents, not just the availability of local jobs, can significantly impact on the employment rate of residents.
The rest of the section focuses on the employment rate data from Table 2. It focuses on those subregions which have either public or private sector employment rates above or below the UK average. Only those results that are statistically significant at the 95% level are included1.
There are 15 subregions that have private sector employment rates that are statistically significantly higher than the UK average and 11 that have rates that are statistically significantly lower than the UK average. For the remaining 11 subregions, the private sector employment rates cannot be shown to be statistically different from the UK average.
Figure 4 shows those subregions with private sector employment rates above the UK average. The highest employment rates are seen in the subregions of North Eastern Scotland and Surrey East and West Sussex. The subregions shown in Figure 4 are spread across England and Scotland with a number of subregions in the north and midlands of England included in the list as well as subregions in the south of England.
Figure 5 shows those subregions with private sector employment rates below the UK average. The lowest employment rates are seen in the NUTS 2 subregions of West Wales and the Valleys, and Merseyside. Northern Ireland, the north east of England and the NUTS 2 subregion of West Midlands are also shown to have relatively low private sector employment rates.
Compared to the private sector, employment rates in the public sector show less variability. For 15 of the 37 subregions, the public sector employment rate of residents can not be shown to be statistically different from the UK average. The remaining 22 subregions are shown in Figures 6 and 7.
Figure 6 shows the 9 subregions that have a public sector employment rate below the UK average. Both Inner and Outer London are included in the list alongside two other subregions in the southeast of England, Essex, and Surrey East and West Sussex. A number of relatively rural subregions of England are also included as well as the more urban subregions of West Yorkshire and West Midlands. The Nuts 2 sub-region of West Midlands is the only subregion to have both public sector and private sector employment rates below the UK average.
Figure 7 shows the subregions with a public sector employment rate above the UK average. These include three subregions in Scotland, the two subregions of Wales and seven subregions in England. Amongst the English subregions, North Yorkshire has the highest employment rate in the public sector amongst its residents. The highest public sector employment rate overall is to be found in Highlands and Islands in Scotland. This reflects the high service provision needs in remote rural areas and Island communities. There is only one NUTS 2 subregion in the UK that has both public sector and private sector employment rates above the UK average and that is Hampshire and Isle of Wight.
A similar analysis to that carried out for NUTS 2 subregions has been carried out at local authority level for Great Britain (GB). For local authorities the results are shown in this article in map form. Maps are included in the article below, and additionally can be viewed interactively via a set of Google maps. The underlying data are available in the data section.
Maps 1 and 2 show public and private sector employee density respectively while Maps 3 and 4 show the public sector and private sector employment rates of local authority residents.
It needs to be stressed that confidence intervals surrounding the estimates are larger at this smaller geographical level. The maps included here cannot show these confidence intervals, they only show the point estimates for 2010 data. Therefore, care needs to be taken in interpreting the maps as it cannot be assumed, for example, that an area in one range band on the map has a statistically significantly higher rate than an area in the band below.
When looking at these maps it is also worth remembering that population is not evenly spread and that as such local authorities in more rural areas are generally larger geographically. When possible, it is often worth zooming in to examine the maps at a more localised level in order to get a fuller picture of patterns in urban areas. By clicking on the links below the maps it is possible to view a larger version. The Google maps provide another method to zoom in on local area data.
With these provisos, Map 1 shows public sector employee density by local authority and Map 2 shows private sector employee density. A number of factors can influence employee density and this includes the boundaries for any particular local authority. A local authority that is largely residential is likely to have a low employee density, while a local authority that includes a major centre of jobs is likely to have a higher employee density. This is seen most clearly in Map 2 where the location of London’s private sector employee jobs can be seen concentrated in the inner core of the city and to the west of London around Heathrow, while the largely residential remainder of Inner and Outer London have employee densities that are low.
Map 3 shows the public sector employment rate by local authority. Public sector employment rates are generally above the GB average across most local authorities in Scotland and Wales. In Scotland, they are highest in the Highlands and Islands, where public sector employment is higher due to the service provision needs of remote rural areas/island communities.
Across England, there is a mix of both high and low public sector employment rates. Furthermore, the local authorities with either high or low rates are relatively spread across different areas of the country. In other words, within England, there doesn’t appear to be any clear north-south divide in terms of public sector employment rates. There are plenty of local authorities in the southern half of England with above average public sector employment rates and plenty of local authorities in the northern half of England with below average public sector employment rates.
By contrast, Map 4 does appear to show clear signs of clustering of high (and low) private sector employment rates within England. Many of the local authorities with above average private sector employment are found clustered in the south of the country, particularly to the north and west of London. Clusters of local authorities with below average private sector employment rates are more typically found in the Midlands and Northern regions of England. More information on the relative clustering of public and private sector employment rates can be found in the next section (Results and Analysis – A Statistical Measure of Clustering).
Scotland, meanwhile, has a wide mix of private sector employment rates. In Aberdeenshire the private sector employment rate is very high. In Glasgow it is low. In Wales, meanwhile, private sector employment rates are below the GB average across nearly all of its local authorities.
As stated above the maps show the point estimates for each subregion but do not take into consideration the confidence intervals surrounding each data point. In compiling Table 3, by contrast, the confidence intervals have been considered and the public sector and private sector employment rate estimates for each local authority have been tested to determine whether they are statistically different from the UK average.Table 3 lists those local authorities which had both public sector and private sector employment rates that were statistically different to the UK average in 2010.
|Private sector employment rate above UK average and public sector employment rate below UK average||Both private and public sector employment rates above UK average||Private sector employment rate below UK average and public sector employment rate above UK average||Both private and public sector employment rates below UK average|
|Elmbridge||Anglesey, Isle of|
|Hammersmith and Fulham||Cardiff|
|Herefordshire, County of||Ceredigion|
|North Lincolnshire||Merthyr Tydfil|
|Pendle||Rhondda, Cynon, Taff|
|Thurrock||Vale of Glamorgan, The|
|Tunbridge Wells||Argyll & Bute|
|Windsor and Maidenhead||East Renfrewshire|
|Wycombe||Eilean Siar (Western Isles)|
In addition to the local authorities listed in table 3 there were a further 63 which had private sector employment rates above the UK average (with public sector employment rates not statistically different to the UK average) and 36 with private sector employment rates below the UK average (with public sector employment rate not statistically different to the UK average).
Additionally there were 26 local authorities with public sector employment rates above the UK average and 17 with public sector employment rates below the UK average (in both cases with a private sector employment rate not statistically different to the UK average). A large number of local authorities (169 out of 379) had neither public nor private sector employment rates statistically different to the UK average. Further details are available in the accompanying data tables.
The previous section included analysis based on visual inspection of the maps. It is possible to supplement this with a statistical analysis that measures the degree of clustering of a variable. This attempts to answer the question:
is the spatial distribution random - in which case levels of public sector employment in a local authority would be entirely independent of the levels in neighbouring authorities, or,
is the spatial distribution clustered - such that a local authority with high levels of public sector employment is likely to be surrounded by other local authorities with similarly high levels of public sector employment (thereby forming a cluster).
This can be answered using a statistical technique known as the Moran I statistic. The Moran I statistic tests for spatial (auto)correlation among neighbouring areas. A value of 0 for the Moran I statistic indicates a random pattern in the spatial distribution of a variable. A value above 0 indicates the existence of spatial clustering with the amount of clustering increasing as the value of the statistic rises from 0 towards 1.
|Private sector employment rate||0.452|
|Public sector employment rate||0.286|
Table 4 presents the Moran I statistic for public sector employment rates and for private sector employment rates in Great Britain. The results show that spatial clustering is more prevalent among private sector employment rates than public sector employment rates.
This is not a surprising result. Many jobs in the public sector are in health, education or public administration and these jobs are likely to be spread relatively evenly across the country in order to serve local populations. By contrast, there are a number of private sector occupations in which firms are not required to locate adjacent to their customers and this gives them a freedom to choose where in Britain (or elsewhere) they wish to locate. For some private sector industries, this results in some geographical clustering (for example the clustering of financial sector activities in Central London).
The Moran I statistic results also support what can be seen visually in Maps 3 and 4 in the previous section. Map 4 showed clustering of high private sector employment rates particularly to the immediate west and north of London and some clustering of lower private sector employment rates elsewhere in England, Scotland and Wales. In Map 3, there is clearly some clustering; for example there are above average public sector employment rates across much of Scotland and Wales. However, across England in particular, local authorities with either a high or low public sector employment rate appear to be more likely to be distributed randomly across space than to be clustered together.
|Private sector employment rate||0.384|
|Public sector employment rate||0.241|
This latter point is confirmed by Table 5. This gives the Moran I statistic results for England only. It shows a Moran I statistic for public sector employment in England of 0.241 which is lower than the 0.286 for the UK. A lower value suggests a more random spatial distribution. Meanwhile, as was the case for Great Britain, within England there is a higher level of clustering of private sector employment than public sector employment.
You may use or re-use this information (not including logos) free of charge in any format or medium, under the terms of the Open Government Licence, write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: firstname.lastname@example.org.
Richard Prothero, Regional Economic Analysis
Tel: +44 (0)1329 447825
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: email@example.com