Uneven Development in the UK Computer Software and Services Industry, 1995-2003: Benchmarking Northern Ireland
Prepared for Invest Northern Ireland by Mike Crone Economic Research Institute of Northern Ireland
Final Report: July 2005
Copyright ©
2005 ERINI
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic chemical, mechanical, optical, photocopying, recording or otherwise without prior permission in writing from the copyright owner.
Published by: Economic Research Institute of Northern Ireland Floral Buildings, 2-14 East Bridge Street Belfast BT1 3NQ Tel: 028 90727350
Contents
Contents.................................................................................................... i Preface..................................................................................................... ii Acknowledgements.....................................................................................iii Executive Summary ................................................................................... iv 1.
Introduction, Definitions and Data Sources ............................................... 1 1.1 1.2
Introduction........................................................................................................ 1 Defining the Sector............................................................................................. 1
1.3
Data Sources...................................................................................................... 4
2. The National Context: Employment in Division 72 and the Wider ICT Sector, 1995-2003 ................................................................................................ 8 2.1 2.2
UK Division 72 Employment ................................................................................ 8 Employment in the Wider ICT Sector................................................................. 10
3. The Spatial Distribution of Division 72 Employment at Regional, County and Local Scales............................................................................................. 18
4.
3.1
Overview .......................................................................................................... 18
3.2 3.3
The Regional Scale ............................................................................................ 18 The County Scale .............................................................................................. 25
3.4
The Local Scale ................................................................................................. 28
Relative Specialisation: Location Quotient Analysis of Division 72 Employment .. ...................................................................................................... 34 4.1
The Location Quotient....................................................................................... 34
4.2 4.3
Regional Location Quotients.............................................................................. 35 County Location Quotients ................................................................................ 38
4.4
UALAD Location Quotients................................................................................. 41
5. Recent Changes in Division 72 Employment: Investigating the Spatial Impact of the IT Downturn ................................................................................... 48 5.1 5.2
Overview .......................................................................................................... 48 Recent Changes at the Regional Scale ............................................................... 48
5.3
Recent Changes at the County Scale ................................................................. 52
5.4
Recent Changes at the Local Scale .................................................................... 56
6. Analysis of Recent Foreign Direct Investment Patterns and Trends in the Software Sector ....................................................................................... 58 6.1 6.2
Overview .......................................................................................................... 58 Software FDI Patterns and Trends, 2000-2004 .................................................. 58
6.3
Comparison with Software FDI Patterns, 1996-99.............................................. 61
References .............................................................................................. 64 Appendix A: NASDAQ Stock Exchange Indices, 1997-2004 .............................. 66 Appendix B: County Map of Great Britain ...................................................... 67 Appendix Ci: Unitary & Local Authority District Map of the UK........................... 68 Appendix Cii: Local Authority District Map of Scotland ..................................... 70 Appendix Ciii: Map of Northern Ireland NUTSIII regions .................................. 71 Appendix D: Location Map of British Provincial Cities....................................... 72 Appendix E: Membership of the ‘Western Arc’ in 2003 (& 2001) ........................ 73
i
Preface
Over the last decade, the information technology industries – including computer software and services – have been among the leading growth sectors of the UK, Irish and global economies, making them the subject of intense policy interest. In two previous reports published by the Northern Ireland Economic Research Centre (Crone, 2000 and 2001) official data sources were used to profile the distribution of employment in computer software and services across the UK regions and Ireland in the periods 1991-97 and 1995-99. Since these earlier reports there has been a period of global uncertainty in the IT sector, manifested in a downturn in demand for IT goods and services and turmoil on global stock markets. This new report seeks to examine the impact of these changed circumstances on employment in the comp uter software and services sector. The report uses official employment data from the ONS Annual Business Inquiry, and equivalent data for Northern Ireland and the Republic of Ireland, to provide a descriptive account of employment patterns and trends since 1995 at three spatial scales (regional, county and local). In particular, the report seeks to explore the spatial employment impacts of the turmoil in the global IT industry by focusing on changes in the period 1999-2003 (i.e. the period since the last NIERC report). The report also includes a simple analysis of recent software inward investment patterns and trends, using data from the Invest UK news archive. The primary aim of the report is to enable local policy-makers, and other key local actors, to benchmark Northern Ireland’s recent employment performance in this key economic sector. The report also hopes to provide an empirical grounding for ongoing debates about the sector’s role and importance within the Northern Ireland economy. Readers of this report should also be aware of a forthcoming ERINI Working Paper by the author and Dr Neil M. Coe (University of Manchester) which seeks to offer some explanatory insights into the patterns of uneven development highlighted in this largely descriptive report #.
#
Coe N M, Crone M (2005, forthcoming) Entrenched uneven development: exploring the (un)changing geography of the UK software and computer services sector since 1995. ERINI Working Paper Series.
ii
Acknowledgements
This report was prepared with financial support from Invest Northern Ireland’s ICT Directorate. The report is part of ERINI’s ongoing research programme on Tradeable
Services.
Excellent
research
assistance
was
provided
by
Johan
Casparsson and Stephen McConville (ERINI). Any errors or omissions are the sole responsibility of the author. Any opinions expressed in the report are those of the author and not those of ERINI or Invest NI.
iii
Executive Summary
Employment in Division 72 and the Wider ICT Sector (Section 2) 1. In December 2003, Division 72 (computer software and services) employment for the UK as a whole stood at 491,400 (approximately 1.9% of all employee jobs). Division 72 employment experienced particularly high growth rates (over 15% per annum) between 1995 and 1999 then growth slowed slightly in 2000 and fell to only 2.5% during 2001. 2002 was a watershed year for the sector with total UK employment falling (for the first time in over a decade) by around 3.4%. The decline in employment was short-lived, however, as the UK as a whole experienced a marginal increase in Division 72 employment during 2003. 2. In December 2003, there were 1.157 million employee jobs in the wider ICT sector in Great Britain (4.5% of all employee jobs). Employment increased by over 5% during both 1999 and 2000, before decreasing by around 1% in 2001, and around 4% during both 2002 and 2003.
3. The ICT sector, can be divided into four broad sub-sectors: i.
Computer software and services (i.e. Division 72): 42% of GB ICT employees in December 2003;
ii.
Wholesale of ICT products (21% of GB ICT employees);
iii. Telecommunications (21% of GB ICT employees); iv. ICT manufacturing (16% of GB ICT employees). The ‘downturn’ had a differential impact on employment in these sub-sectors. ICT wholesale employment, and particularly ICT manufacturing employment, declined significantly between 1998 and 2003, and at a faster rate after 2000. In
contrast,
telecommunications
and
computer
software
and
services
employment were increasing rapidly until 2001 and 2002 respectively and experienced only slight declines thereafter.
4. Comparison of the NI and GB ICT sectors in 2001 and 2003 shows that NI was more specialised in ICT manufacturing employment and less specialised in computer software and services. Between 2001 and 2003, the rate of decline of ICT manufacturing employment was the same in NI and GB but computer
iv
software and services employment declined more severely in NI than in GB. Overall, NI’s recent ICT sector employment performance was relatively poor.
The Spatial Distribution of Division 72 Employment (Section 3) 5. At the regional scale, the spatial distribution of Division 72 employee jobs is highly uneven. In 2003, London and the South East – which had 97,600 and 123,900 employees respectively – together accounted for just under half of all GB computer software and services employees. And two further regions – East and North West, with around 50,000 employees each – accounted for a further fifth. The remaining seven GB regions accounted for only one third of GB computer software and services employees. The ‘Greater South East’ (i.e. London, South East and East) consistently accounted for around 60% of GB Division 72 employees until 2001 after which it fell slightly. All the GB regions experienced strong net employment growth in Division 72 during the period 1995-2003. However, the average growth rate was lowest in London and East of England and highest in North East and Scotland (North West and Yorkshire and the Humber also had growth rates above the GB average).
6. According to official data, the number of Division72 employees in Northern Ireland increased rapidly from only 1,380 in September 1995 to a peak of 5,780 in September 2001. It then declined to around 5,100 in June 2003 (a fall of 11.4% from the peak) before recovering slightly to 5,360 in March 2004. ERINI PSU estimates put the December 2002 figure closer to 7,000. NI was the fastest growing UK region in terms of Division 72 employees between 1995 and 2003. One interpretation is that NI was beginning to ‘catch-up’ with other GB regions from a low starting point. 7. Division 72 employment in the Republic of Ireland grew from only 5,380 in 1995 to
over
19,680
in
2002 (2003 data
not
yet
available).
Heading
223
“Reproduction of recorded media” (covering the reproduction of packaged software) employed 6,000 further persons in 2002 (up from 4,000 in 1995). ROI Division 72 employment grew at a faster rate than all UK regions except NI during 1995-2002.
8. The county distribution of Division 72 employees is even more uneven: the top third of counties covered 78.4% of GB employees in 2003. The Top 10 counties
v
in 2003 comprised London, five ‘non- metropolitan’ counties located in a contiguous ‘arc’ to the western side of London (Berkshire, Surrey, Hampshire, Hertfordshire and Buckinghamshire), three ‘metropolitan’ counties located further north (West Midlands, Greater Manchester and West Yorkshire) and Avon. Berkshire had 35,000 computer software and services employees in 2003 (more than NI and ROI combined!). Greater Manchester, the leading northern county, had over 21,000 computer software and services employees.
9. Belfast is the dominant location of computer software and services jobs in NI, with 2,900 employee jobs in September 2003 (55.5% of NI total). For comparison, note that seven of the leading British provincial cities had 5,000 or more
Division
72
employees
in
2003
(Leeds,
Birmingham,
Edinburgh,
Nottingham, Manchester, Glasgow and Bristol). However, Belfast did have more Division 72 employees than several other comparable GB cities.
Division 72 Location Quotients (Section 4)
10. In 2003, regional LQs ranged from 183 (South East) to 39 (Wales), indicating considerable inter-regional variation in the degree of employment specialisation in Division 72. Four distinct ‘tiers’ can be identified in the 2003 regional LQ hierarchy. The first tier comprised the three regions of the Greater South East (LQ>100). The second tier (LQ 80-90) comprised three regions stretching down the western side of England: North West, West Midlands and South West. The third tier (LQ 60-75) comprised four regions in the northeastern part of the UK: East Midlands, Yorkshire and the Humber, North East and Scotland. The final tier (LQ100) in 2003 were two clusters of local government areas: one around the southern fringe of Manchester ; the other on the southern and eastern fringe of Birmingham. 13. Belfast had a 2003 LQ of 82, and was ranked 156th out of 413 UK UALADs. Belfast was much better placed in the computer software and services sector, in both absolute and relative terms, than Aberdeen, Swansea, Liverpool, Hull, Bradford, Cardiff and Leicester. However, Belfast trailed well behind the cities of Nottingham, Bristol, Edinburgh, Newcastle-upon-Tyne, Leeds and Manchester in both absolute and relative terms. The other four NI NUTSIII regions compared much less favourably than Belfast, having 2003 LQs ranging from 49 in the North of NI to only 7 in the West & South of NI (below all 408 GB local government areas).
The Spatial Impact of the IT Downturn (Section 5)
14. Nationally, Division 72 employment grew every year from 1995 to 2001 before declining during 2001-02. Focusing on the three years after December 2000, we find that the downturn had a regionally differentiated impact on computer software and services employment. More specifically, employment: ⇒ declined in London in all three years (2000-01, 2001-02 and 2002-03); ⇒ declined in 2001-02 and 2002-03 in South East, Scotland and Northern Ireland (with the more severe decline in 2001-02 in all cases); ⇒ declined in 2001-02 but returned to growth in 2002-03 in South West, East Midlands and Wales; ⇒ declined in 2000-01 but returned to growth thereafter in East; ⇒ continued to expand in all three years (but at a slower rate than previously) in the North West and Yorkshire and the Humber; ⇒ and continued to expand in 2000-01 and 2001-02, with only a marginal decline in 2002-03, in West Midlands.
vii
15. Analysis at a finer geographical scale shows that the Western Arc was disproportionately impacted by the IT downturn. Collectively, computer software and services employment in the 37 members of the Western Arc declined by 9.6% (nearly 13,000 jobs) in the period 2001-03, compared to 3.2% in the UK as a whole. The downturn had a highly variable impact on the leading provincial cities outside the GSE during 2001-03. Thus, computer software and services employment continued to expand (to varying degrees) in Leeds, Sheffield, Nottingham, Manchester and Bristol but declined quite sharply (by more than 10%) in Birmingham, Edinburgh and Glasgow. Belfast had a more neutral experience
with
only
a
marginal
(negative)
change
in
its
Division
72
employment between 2001 and 2003. However, the remainder of NI (i.e. all district council areas outside Belfast) was much more adversely affected, losing 17% of its Division 72 employees between 2001 and 2003
Recent Software FDI Patterns and Trends (Section 6)
16. A total of 115 software FDI projects were attracted to the UK during 2000-2004. New developments were the dominant type of project during the period, accounting for 71% of all projects. There were also 20 acquisitions, 11 expansion projects and 2 joint ventures. The United States was the major source of software FDI projects into the UK during 2000-04, accounting for 78% of all projects. Canada (10 projects) was the other major source country. Firms from Europe and from the rest of the world each accounted for 11 software FDI projects in this period. London (40 projects), the South East (30) and East of England (10) attracted the largest number of software FDI projects. Thus FDI in this period was acting to reinforce to existing dominance of the UK computer software and services industry by the Greater South East. 17. There was a dramatic shift in the regional location behaviour of foreign investors between the two periods. The Greater South East accounted for only 38% of all projects attracted to the UK during 1996-99 but this increased to 70% during 2000-04. In contrast, the number of projects attracted to the rest of the UK fell in the latter period. For example, during 1996-99, Scotland and NI attracted 17 and 16 projects respectively, but during 2000-04 they were only able to attract 9 and 8 projects respectively.
viii
1. Introduction, Definitions and Data Sources
1.1 Introduction a. The primary aim of the report is to enable local policy-makers, and other key local actors, to benchmark Northern Ireland’s recent employment performance in the “computer software and services sector” against other regions and sub-regions of the British Isles. The report hopes to provide an empirical grounding for ongoing debates about this key sector’s role and importance within the Northern Ireland economy. b. The report is divided into five substantive sections. The first section looks at national employment trends in computer software and services, and in the wider ICT sector. The second section uses official employment data to provide a descriptive account of employment patterns and trends in computer software and services since 1995 at three spatial scales (regional, county and local). The third includes a location quotient analysis of relative employment specialisation in computer software and services, also at three spatial scales. The fourth section of the report explores the spatial employment impacts of the recent turmoil in the global IT industry, focusing on the period 2000-2003. The fifth section consists of a simple analysis of recent inward investment patterns and trends in computer software and services.
1.2 Defining the Sector c. Those parts of the economy engaged in the production and delivery of what might be loosely termed “information and communication technology products
and
governments
services” and
have
economic
become
a
development
key
focus
agencies
of
attention
around
the
for
world.
However, analysis and discussion of this industry (or, perhaps more accurately, group of industries) has been complicated by its blurred boundaries
(a
technologies),
reflection by
the
of
evolving,
sloppy,
and
1
and
sometimes
sometimes
converging,
inconsistent,
use
of
terminology, and by the inflexibility and infrequent modification of official industrial classification systems. d. Thus, efforts by researchers to quantify or measure the industry are highly contingent upon the approach employed. This often boils down to two key decisions: first, defining the boundaries of the industry (or a part of it) and, second, choosing between official data sources (less costly, more readily available, possibly more reliable but often less than ideal for the purpose) and bespoke primary data collection (more costly, more difficult, sometimes less reliable but potentially more fit for the purpose).
e. This report relies on official statistics and is therefore bound by the definitions imposed by the 1992 Standard Industrial Classification (and its most recent 2003 revision). Specifically, the majority of the analysis is confined to employment data for Division 72 ‘Computer and related activities’1 . Division 72 is an aggregation of a number of different activities including software consultancy and supply, hardware consultancy, data processing, database activities, computer maintenance and other computer related activities (see Table 1.1 for a full definition). Thus the sector approximates to what we might otherwise call the “computer software and services” industry. Readers should be aware that in official employment statistics each separate business is classified according to its primary business activity. This means that employment in an IT department in a bank or financial institution, for example, would not be counted in the official figure for Division 72. Division 72 employment therefore refers only to employment in businesses whose primary business activity is one of the activities outlined in Table 1.1. f.
Clearly, Division 72 is a subset of the wider group of industries typically referred to as ICT (Information and Communications Technology). “ICT” has become common currency in government circles but was initially poorly defined. However, in 2001, the Office for National Statistics and the Organisation for Economic Co-operation and Development (OECD) issued an agreed definition of the ICT sector, based on a grouping of various divisions and sub-headings of the 1992 Standard Industrial Classification (shown later in Table 2.2). Using this definition, ICT is a considerably more broad-based group
1
of
industries
than
Division
72,
including
the
IT
hardware
The exception is Section 2, which also includes an analysis of the wider ICT sector – see below.
2
manufacturing and telecommunications industries, and the wholesale of ICT products, in addition to computer software and services. To help set Britain’s Division 72 (“computer software and services”) sector within the wider context of the ICT sector, therefore, Section 2 of this report presents employment data for the wider ICT sector and its various sub-sectors (following the official definition).
Table 1.1: Definition of Division 72 ‘Computer and Related Activities’
72.10
Hardware consultancy
72.20
Software consultancy and supply This class includes: analysis, design and programming of systems ready to use: § analysis of user’s needs and problems, consultancy on best solution § development, production, supply and documentation of order-make software based on orders from specific users § development, production, supply and documentation of ready-made (noncustomised) software § writing of programs following directives of the user This class excludes: reproduction of non-customised software cf. 22.33 software consultancy related to hardware consultancy cf. 72.10
72.30
Data processing This class excludes: processing of data employing either the customer’s or a proprietary program: § complete processing of data § data entry services management and operation on a continuing basis of data processing facilities belonging to others
72.40
Data base activities This class includes data base related activities: data base development: assembly of data from one or more sources data storage: preparation of a computer record for such information in a predetermined format data base availability: provision of data in a certain order or sequence, by on-line data retrieval or accessibility (computerised management) to everybody or to limited users, sorted on demand
72.50
Maintenance and repair of office, accounting and computing machinery
72.60
Other computer related activities
Source: UK Standard Industrial Classification of Economic Activities 1992, ONS.
3
1.3 Data Sources
1.3.1 The ONS Annual Business Inquiry for Great Britain
g. The Office for National Statistics’ Annual Business Inquiry (ABI) has been the official source of annual employment data for Great Britain since 1998, when it replaced the Annual Employment Survey (itself a replacement for the earlier biennial Census of Employment). The ABI is a sample survey that records the number of employees in employment in December of each year. At the time of writing, the most recently released data from the ABI refer to December 2003. Thus, a continuous series of annual employment data are available for the period December 1995 to December 2003 2 .
h. ABI data can be obtained in a sectorally and spatially disaggregated form (via NOMIS). This allows us to analyse employees in employment data for Division 72 at three spatial scales. Regional data are available for the nine English Government Office Regions (i.e. the scale at which English Regional Development Agencies operate) and for Scotland and Wales. Data are also available for current local government areas (i.e. the 408 GB Unitary Authorities and Local Authority Districts). For an intermediate sub-regional scale, we can obtain data for the English and Welsh counties and Scottish regions that existed prior to the 1996 reorganisation of local government boundaries (although data are no longer available at this scale from 2003).
1.3.2 Northern Ireland Data
i.
Northern Ireland (hereafter NI) is not covered by the ONS Annual Business Inquiry, instead having its own statistical series for measuring employment at the workplace level: the biennial Census of Employment and Quarterly Employment Survey. These data series offer broad comparability with the GB ABI and are used in this report to benchmark NI’s Division 72 employees
2
Johnston (2001) revealed that total employment in GB under the 1998 ABI was revised up by approximately one million (4.2 per cent) on the 1998 AES figure, from just under 24 million to just under 25 million. Because of this discrepancy, which ONS attributes to the elimination of undercounting under the old methodology, all AES data from 1995-97 were “re-scaled” to be consistent with subsequent ABI data. For a detailed examination of the implications of the change from AES to ABI for Division 72 employees data see Crone (2001), pages 1-6.
4
in employment against the GB regions and counties. The data used in the report were supplied to the author by DETI Statistics Research Branch. Note that data for September 1995, 1997, 1999, 2001 and 2003 come from the NI Census of Employment and annual data for inter-Census years come from the September Quarterly Employment Survey 3 . j.
Recent research by ERINI’s Priority Skills Unit (commissioned by the NI Skills Task Force to investigate the labour market for IT skills) has raised doubts over the accuracy of official figures for Division 72 employees in employment in NI. Due to the possibility of significant undercounting in the official data (McGuinness and Doyle, 2003), ERINI PSU employment estimates are also included in the subsequent analysis.
k. Unfortunately there is no published sub-regional (e.g. county or district council) analysis of CoE or QES data at two-digit SIC level (i.e. for Division 72). In particular, disaggregation of two-digit employment data to district council level raises confidentiality constraints due to the small size of the 26 NI district council areas. However, to allow for a limited comparison of NI sub-regions with GB counties and local authorities, a special analysis of twodigit employment data from the 2001 and 2003 NI CoE was obtained from DETI Statistics Research Branch. The sub-regional scale utilised for this analysis is the European NUTS III region, of which there are five in NI (Belfast, Outer Belfast, East of NI, North of NI, and West & South of NI). This is the smallest geographic scale at which Division 72 employment data can be published without breaching confidentiality constraints.
1.3.3 Republic of Ireland Data l.
Since the Central Statistics Office publishes data according to the European standard for industrial classification (NACE Revision 1) it is possible to make a direct comparison of Division 72 employment in the Republic with that in NI and the GB regions and sub-regions. Specifically, data from the CSO’s Annual Services Inquiry (ASI) are broadly comparable to the GB Annual Business Inquiry and NI QES. The ASI has surveyed the “Business Services” sectors (which include Division 72) on an annual basis since 1995
3
It should be noted here that, whilst GB ABI data refer to employees in employment in December of each year, the NI Census of Employment data refers September. Therefore, September QES data are
5
(previously it was on a three year rotating basis). At the time of writing, the most current ASI data refer to 2002, giving a continuous series for 19952002. This limits our ability to compare the impact of the downturn in the IT sector on Ireland with that on the UK regions and sub-regions.
m. Another issue that must be addressed when comparing employment in the computer software and services sector in Ireland with other regions is the presence in Ireland of a particular sub-sector of the software industry that engages in the localisation, manufacturing and distribution of mass- market packaged software (Coe, 1997; Crone, 2002a) 4 . This sub-sector accounts for a significant portion of the “Irish software industry” but is more or less unique to Ireland within Europe (Crone, 2001). The particular issue here is that this sub-sector is classified under Heading 2233 of the NACE industrial classification system (“Reproduction of Computer Media”) rather than Division 72 5 . Employment data for this sub-sector in Ireland, which are considered in Chapter 3 of this report, are obtained from the CSO’s Census of Industrial Production.
1.3.4 Inward Investment Data
n. The final substantive chapter of this report looks
at
recent
inward
investment patterns and trends in the computer software and services sector. This chapter is based on an analysis of articles from the searchable news archive on the UK Trade & Investment web-site (www.investuk.com). UK Trade & Investment is the UK Government’s inward investment agency, covering the whole of the United Kingdom. The news articles on the InvestUK web-site include details of the date of investment, investment country of origin, region (i.e. location of investment) and industry sector6 . A search of the archive for “software” sector investments from the start of the used for consistency with the NI Census. 4
Firms in this sub-sector are engaged in activities such as language translation, CD-ROM manufacturing and associated packaging, distribution and fulfilment. They include the Irish subsidiaries of US software multinationals (e.g. Microsoft, Oracle and Symantec) and specialist sub-contractors who work for the software multinationals in Ireland (e.g. Banta Global Turnkey, Modus Media, Saturn Fulfilment, Sonopress and Zomax) – most of which are multinationals themselves. 5
The presence of Heading 2233 accounts for the majority of the observed discrepancy between the National Software Directorate’s estimate of the size of the Irish software industry and the official employment data contained in the ASI under Division 72 (Crone, 2001). 6
Note that the events reported in the news archive are foreign direct investments. That is, investments in the UK by non-UK companies. Inter-regional investments by UK-owned companies are not covered.
6
archive in January 2000 until the end of May 2004 was conducted in June 2004, revealing 115 different projects. Summary tables were compiled categorising these projects by year, country of origin, type of investment and location of investment. A similar dataset covering the period April 1996 to
September
1999,
compiled
by
Crone
(2001),
was
available
for
comparative purposes.
o. One shortcoming of this dataset is the absence of reliable information on either the number of employee jobs created by each investment or the monetary value of the investment. This means that any analysis is limited to numbers of projects, with no means of differentiating between projects of different size. Thus a project creating 100 jobs is given the same weight as a project creating 5 jobs, for example. This limitation should be born in mind when digesting the analysis in Chapter 6. Nevertheless, the available dataset offers some useful insights into the geographical distribution and character of recent software inward investment projects.
From a Northern Ireland perspective, this excludes investments by companies from the GB mainland (which are treated as inward investors by Invest Northern Ireland).
7
2. The National Context: Employment in Division 72 and the Wider ICT Sector, 1995-2003
2.1 UK Division 72 Employment a. In December 2003, Division 72 employees in employment for the UK as a whole stood at approximately 491,400 (Table 2.1). This represents approximately 1.9% of all employee jobs in the UK. b. The number of Division 72 employees more than doubled in the eight year period 1995-2003; equivalent to a compound annual growth rate (CAGR) of 9.5%. c. Impressive employment growth has been a hallmark of the UK computer software and services sector since the early 1980s. For example, employment in Activity Heading 8394 of the 1980 Standard Industrial Classification (the predecessor of Division 72) increased from 54,800 in 1981 to 147,500 in 1991 (CAGR 10.4%) (Coe, 1996a). Then, in the early 1990s (1991-95) Division 72 employment increased from 170,400 to 237,800 (CAGR 8.7%) d. The period 1995-2001 was another period of sustained growth for the sector, with particularly high growth rates (over 15% per annum) between 1995 and 1999 (Table 2.1 and Figure 2.1). Growth slowed slightly in 2000 and fell to only 2.5% during 2001. The year 2002 was a watershed year for the sector with total UK employment falling for the first time in over a decade; the number of employee jobs fell by around 3.4% from the 2001 peak of 507,500 (Table 2.1). The decline in employment was short-lived, however, as the UK as a whole experienced a marginal increase in Division 72 employment during 2003 7 . As Figure 2.1 shows, these recent UK Division 72 employment trends seem to follow international market fortunes in the technology sectors, as proxied by the performance of the Nasdaq stock exchange (see Appendix A). Specifically, the rapid growth, slower growth, slight retrenchment and tentative recovery phases of the past eight years have followed (with a time lag) the boom, crash, steady decline and slight
7
Note though that in Dec 2003 UK total employment in Division 72 had not yet recovered to Dec 2000 levels.
8
recovery phases of the Nasdaq index. Thus, we have prima facie evidence of the impact of global tech sector downturn on UK Division 72 employment. The regional and sub-regional impacts of this phenomenon are explored in Chapter 5.
Table 2.1: GB and UK Employees in Employment in Division 72, 1995-2003
GB NI UK GBYoY %chg NIYoY %chg UKYoY %chg
1995
1996
1997
1998
1999
2000
2001
2002
2003
236,459
272,075
314,359
374,932
431,231
490,003
501,748
484,704
486,177
1,380
1,680
2,240
3,420
3,970
4,940
5,780
5,470
5,247
237,839
273,755
316,599
378,352
435,201
494,943
507,528
490,174
491,422
NA
15.1%
15.5%
19.3%
15.0%
13.6%
2.4%
-3.4%
0.3%
NA
21.7%
33.3%
52.7%
16.1%
24.4%
17.0%
-5.4%
-4.1%
NA
15.1%
15.7%
19.5%
15.0%
13.7%
2.5%
-3.4%
0.3%
Notes: YoY %chg = year-on-year percentage change. Percentage increases over 1995-2003: GB=105.6%; NI=280.2%; UK=106.6%. Equivalent compound annual growth rates: GB=9.4%; NI=18.2%; UK=9.5%. Sources: GB data from ONS Annual Business Inquiry 1995-2003 (via NOMIS). NI data from (September) Quarterly Employment Survey & NI Census of Employment, supplied by DETI Statistics Research Branch.
Figure 2.1: Division 72 Employees in Employment, 1995-2003 Peak of NASDAQ
600,000
Boom period for NASDAQ
Great Britain
Crash of NASDAQ
Slower decline of NASDAQ
Recovery?
United Kingdom 500,000
400,000
300,000
200,000
100,000
0 Dec 1995
Dec 1996
Dec 1997
Dec 1998
Dec 1999
Dec 2000
Dec 2001
Dec 2002
Dec 2003
Sources: GB data from ONS Annual Business Inquiry 1995-2003 (via NOMIS). NI data from (September) Quarterly Employment Survey & NI Census of Employment, supplied by DETI Statistics Research Branch.
9
e. Division 72 has become an increasingly important sector of the UK economy, as we can see from an examination of sectoral changes in employment in the period December 1995 to December 2003. Despite a slowing of growth in recent years, the number of Division 72 employees still increased by 106.6% over this eight year period (+253,600 jobs, CAGR 9.5%). This compares, for example, with a: ⇒ 13.2% increase in total UK employee jobs over the same period; ⇒ 18.8% decrease in UK manufacturing jobs (Section D); ⇒ 17.0% increase in UK wholesale and retail jobs (Section G); ⇒ 5.8% increase in UK financial intermediation jobs (Section J); and ⇒ 33.5% increase in UK business services jobs (Section K; includes Div 72). In fact, as a result of its rapid growth, Division 72 accounted for 8.3% of all new employee jobs added in the UK between 1995 and 2003, despite comprising only 1.0% of all UK employee jobs at the start of the period.
2.2 Employment in the Wider ICT Sector f.
The analysis in this section is initially confined to GB due to a lack of equivalent NI data for the four digit sectors comprising the wider ICT sector – although some comparable NI data for 2001 and 2003 are introduced at the end. Based on the OECD definition, there were 1,156,746 employee jobs in the wider ICT sector in GB in December 2003 (Table 2.2); 4. 5% of all employee jobs in GB.
g. Employment in the GB ICT sector increased by over 5% during both 1999 and 2000. It then decreased by around 1% during 2001, and by around 4% during both 2002 and 2003 (Table 2.2). As a result, the number of ICT employees in GB in December 2003 was only marginally greater than in December 1998.
h. The OECD definition divides the ICT sector into a number of sub-sectors. One important distinction is between: ⇒ ICT manufacturing industries – which accounted for 187,800 employee jobs in December 2003 (16.2% of the GB ICT sector); ⇒ ICT service industries – which accounted for 968,900 employee jobs in December 2003 (83.8%% of the GB ICT sector).
10
Table 2.2: GB ICT Sector Employment (as per OECD Definition) by Constituent Sub-sectors, 1998-2003 1998
1999
2000
2001
2002
3001 : Manufacture of office machinery
11,124
10,516
10,265
9,157
8,757
3002 : Manufacture of computers/other information processing equipment 3130 : Manufacture of insulated wire a nd cable
37,705
40,626
40,233
33,719
29,628
19,801
20,274
17,523
15,323
12,400
3210 : Manufacture of electronic valves and tubes and other electronic components 3220 : Manufacture of television and radio transmi tters and apparatus for line telephony and line telegraphy 3230 : Manufacture of television and radio receivers, sound or video recording or reproducing apparatus and associated goods
51,340
50,038
40,133
37,341
31,422
38,223
40,711
54,557
43,342
35,235
30,286
30,381
3320 : Manufacture of instruments and appliances for measuring, checking, testing, navigating and other purposes 3330 : Manufacture of industrial process control equipment
78,573
75,591
10,797
2003
Chg98 -03
%ch98 -03
7,662
-3,462
-31.1%
23,048
-14,657
-38.9%
10,527
-9,274
-46.8%
28,450
-22,890
-44.6%
30,638
25,367
-12,856
-33.6%
28,456
21,729
19,539
-15,696
-44.5%
71,935
74,089
70,288
65,592
-12,981
-16.5%
8,833
8,225
8,554
7,611
7,624
-3,173
-29.4%
282,798
276,875
273,252
249,981
212,473
187,809
-94,989
-33.6%
42,794
44,682
47,908
44,292
38,633
39,891
-2,903
-6.8% -26.7%
ICT MANUFACTURING INDUSTRIES
ICT manufacturing total ICT SERVICE INDUSTRIES 5143 : Wholesale of electrical household goods 5164 : Wholesale: office machinery and equip.
89,139
92,489
85,365
83,249
82,380
65,363
-23,776
5165 : Wholesale: other industry machinery etc
140,473
133,289
135,635
133,307
132,199
132,761
-7,712
-5.5%
Total wholesale of ICT products
272,406
270,460
268,908
260,848
253,212
238,015
-34,391
-12.6%
6420 : Telecommunications
193,357
208,820
236,362
244,053
250,036
241,068
47,711
24.7%
3,072
3,004
2,806
2,901
3,926
3,677
605
19.7%
12,932
17,481
18,877
16,892
17,573
19,283
6,351
49.1%
220,340
247,143
272,196
287,428
281,525
299,885
79,545
36.1%
46,431
40,710
52,609
55,254
43,359
37 ,334
-9,097
-19.6%
6,669
9,193
9,391
10,535
10,527
10,879
4,210
63.1%
7250 : Maintenance and repair of office, c omputing machinery
25,674
27,714
26,682
19,095
28,246
26,898
1,224
4.8%
7260 : Other computer related activities
62,886
88,990
110,248
112,544
103,474
91,898
29,012
46.1%
Total computer software and services
374,932
431,231
490,003
501,748
484,704
486,177
111,245
29.7%
ICT services total
843,767
913,515
998,079
1,009,550
991,878
968,937
125,170
14.8%
1,126,565
1,190,390
1,271,331
1,259,531
1,204,351
1,156,746
30,181
2.7%
7133 : Renting of office machinery and equipment 7210 : Hardware consultancy 7220 : Software consultancy and supply 7230 : Data processing 7240 : Data base activities
ICT SECTOR: TOTAL
Note: Equivalent compound annual growth rate s 1998-2003 for major ICT sub-sectors as follows: ICT manufacturing -5.0%; Wholesale of ICT products -1.7%; Telecommunications +2.8%; Computer software and services +3.3%; Total ICT +0.3%. Source: Data on employee jobs obtained from ONS Annual Business Inquiry (via NOMIS).
11
i.
By further sub-dividing the ICT services sector, we can identify four broad ICT sub-divisions (Table 2.2), as follows: i.
Computer software and services (i.e. Division 72): 486,200 employee jobs in December 2003 (42% of the GB ICT sector);
ii.
Telecommunications: 241, 100 employee jobs in December 2003 (21% of the GB ICT sector);
iii. Wholesale of ICT products: 238,000 employee jobs in December 2003 (21% of the GB ICT sector); iv. And ICT manufacturing (as above): 187,800 employee jobs (16%). j.
An examination of changes in these four main sub-sectors between December 1998 and December 2003 reveals some interesting trends (Table 2.2). In both the manufacturing and wholesale sub-sectors, the number of employee jobs was in decline throughout the period, decreasing by 34% and 13% respectively across the five years. The rate of decline in both sub-sectors was notably more severe after 2000, suggesting the global ICT downturn accentuated an established downward trend (Figure 2.2). In contrast, the telecommunications sub-sector continued to add jobs between 1998 and 2002, albeit at a slower rate after 2000, before contracting by 6% during 2003. The trend in computer software and services was a slowing of the previously rapid employee growth during 2001, a relatively modest decline during 2002 and a slight return to growth during 2003 (as described previously). Thus, the ‘downturn’ seems to have had a differential impact on the various sub-sectors of ICT, with the ICT service industries being notably more resilient than ICT manufacturing, and computer services being the most resilient sub-sector of ICT during the five years since 1998.
k. As a consequence of the changes described above, there was a shift in the relative importance of the various sub-sectors of ICT over the period (Figure 2.3). Thus, ICT manufacturing fell in importance, from a 25% share of total GB ICT employees in 1998 to only 16% in 2003. Similarly, ICT wholesale fell in importance from 24% in 1998 to 21% in 2003. In contrast, telecommunications increased its share of GB ICT employees from 17% to 21%. And computer software and services (Division 72) increased its share of GB ICT employees from 33% in 1998 to 42% in 2003.
12
Figure 2.2: Employees in GB ICT Sub-sectors
Division 72 600,000
Telecommunications
ICT wholesale
Figure 2.3: Share of Employees in GB ICT Sub-sectors
ICT manufacturing
Peak of NASDAQ Nasdaq "boom"
Total computer services
6420 : Telecommunications
Total wholesale
ICT manufacturing total
100%
Nasdaq "crash"
Slower decline
Recovery?
90% 500,000
80%
70% 400,000
60%
300,000
50%
40% 200,000
30%
20% 100,000
10%
0%
0
1998
1999
2000
2001
2002
1998
2003
1999
2000
2001
2002
Source: Analysis of ONS Annual Business Inquiry.
Source: ONS Annual Business Inquiry (via NOMIS).
13
2003
l.
Among the 11 GB regions, there were some notable changes in the size and composition of the wider ICT sector between December 1998 and December 2003 (Figures 2.4 and 2.5). Overall ICT sector employment expanded in six of the 11 GB regions during this period (South East, North West, West Midlands, East Midlands, Yorkshire and the Humber and North East) and contracted in five regions (London, Eastern, Scotland, South West and Wales). The latter group of regions (except London) all suffered from their high relative specialisation in ICT manufacturing. Both the number of jobs and the share of total ICT employment in the ICT manufacturing sub-sector declined in all GB regions during 1998-2003, with the decline being particularly severe in Scotland and Wales (Figures 2.4 and 2.5). In contrast, note that computer services employment expanded (in absolute terms and its share of total ICT) in all 11 GB regions between 1998 and 2003 (Figures 2.4 and 2.5). A final point to note about the regional distribution of ICT sector employment in GB is that the regions with the most ICT jobs are those which are most specialised in computer services, as measured by location quotients (Figure 2.6) – particularly the three regions of the Greater South East but also the North West and West Midlands.
Figure 2.4: Employment in Four Main Sub-sectors of ICT by Region in 1998 & 2003 300,000
250,000
200,000
Total computer services 6420 : Telecommunications
150,000
Total wholesale ICT manufacturing total
100,000
50,000
South East
London
Eastern
North West
West Scotland Midlands
South West
East Yorkshire Midlands
Source: Analysis of ABI employment data obtained from NOMIS.
14
Wales
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
0
North East
Figure 2.5: Share of Total ICT Employment in Four Main Sub-sectors by Region in 1998 & 2003 100%
80%
60% Total computer services 6420 : Telecommunications Total wholesale ICT manufacturing total 40%
20%
South East
London
Eastern
North West
West Scotland Midlands
South West
East Yorkshire Midlands
Wales
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
2003
1998
0%
North East
Source: Analysis of ABI employment data obtained from NOMIS. Note: Regions ranked by total ICT employment in 1998 (as per Figure 2.4).
Figure 2.6: 2003 location quotients for major sub-sectors of ICT by region 200
180
160
140
120 ICT manufacturing total Total wholesale
100
6420 : Telecommunications Total computer services
80
60
40
20
0 South East
London
Eastern
North West
West Midlands
Scotland
South West
East Midlands
Yorkshire
Wales
Source: Analysis of ABI employment data obtained from NOMIS. Note: Regions ranked by total ICT employment in 1998 (as per Figure 2.4).
15
North East
m. Data issues precluded the inclusion of NI in the above cross-regional analysis of the wider ICT sector. However, using data from the 2001 and 2003 NI CoE, we can ma ke a broad cross-sectional comparison of the wider ICT sector in NI and GB, and comment on comparative changes in the intervening two year period (Table 2.3 and Figure 2. 7). First, note the NI ICT sector was relatively specialised in ICT manufacturing, and relatively underrepresented in computer software and services, compared to GB.
n. Second, note the key changes between 2001 and 2003 in GB and NI: ⇒ ICT manufacturing employment declined by 25% in both GB and NI (Table 2.3) but NI lost a greater proportion of its total ICT employment as a result of its greater relative specialisation in ICT manufacturing. ⇒ ICT wholesale employment declined by 9% in GB but increased by 2.4% in NI (Table 2.3); ⇒ Telecommunications employment declined slightly in GB and increased very marginally in NI (Table 2.3); ⇒ Computer software and services declined by 9% in NI compared to only 3% in GB (Table 2.3); ⇒ Overall, the NI ICT sector performed relatively poorly during the period with total ICT employment declining by 11.6% compared to 8.2% in GB.
Figure 2.7: Composition of the ICT Sector in GB and NI in 2001 and 2003 100%
29.4% 80%
39.8%
30.2%
42.0%
19.3%
60%
21.9%
Computer software and services 6420 : Telecommunications
19.4% 20.8%
Wholesale of ICT products
14.1% 40%
16.4%
ICT manufacturing total 20.7% 20.6% 20%
37.2% 31.6% 19.8%
16.2%
0% % of GB 2001
% of GB 2003
% of NI 2001
% of NI 2003
Sources: GB data from ONS Annual Business Inquiry (via NOMIS). NI data from 2001 and 2003 Census of Employment, supplied by DETI Statistics Research Branch.
16
Table 2.3: Wider ICT sector employment - comparison of GB and NI in 2001 and 2003 GB ABI 2001
% of GB 2001
GB ABI 2003
% of GB 2003
NI CoE 2001
% of NI 2001
NI CoE 2003
% of NI 2003
2,466
12.5%
2,325
13.4%
GB ch01-03
GB %ch 01-03
NI ch01-03
NI %ch 01-03
-141
-5.7%
nd
nd
-1,758
-61.4%
75
32.2%
ICT MANUFACTURING INDUSTRIES 3001 : Manufacture of office machinery
9,157
0.7%
7,662
0.7%
3002 : Manufacture of computers etc
33,719
2.7%
23,048
2.0%
3130 : Manufacture of insulated wire and cable
15,323
1.2%
10,527
0.9%
nd
nd
3210 : Manufacture of electronic valves etc
37,341
3.0%
28,450
2.5%
2,068
10.5%
3220 : Manufacture of TV/radio transmitters etc
43,342
3.4%
25,367
2.2%
1,583
8.1%
3230 : Manufacture of TV/radio receivers etc
28,456
2.3%
19,539
1.7%
968
4.9%
3320 : Manuf: instruments for measuring etc
74,089
5.9%
65,592
5.7%
233
1.2%
3330 : Manuf: industrial process control e quip.
-1,495
-16.3%
-10,671
-31.6%
nd
nd
-4,796
-31.3%
-8,891
-23.8%
2,861
16.5%
-17,975
-41 .5%
-8,917
-31.3%
308
1.8%
-8,497
-11.5%
8,554
0.7%
7,624
0.7%
nd
nd
nd
nd
-930
-10.9%
nd
nd
249,981
19.8%
187,809
16.2%
7,318
37.2%
5,494
31.6%
-62,172
-24.9%
-1,824
-24.9%
5143 : Wholesale of electrical household goods
44,292
3.5%
39,891
3.4%
803
4.1%
821
4.7%
-4,401
-9.9%
18
2.2%
5164 : Wholesale: office machinery and equip.
83,249
6.6%
65,363
5.7%
328
1.7%
338
1.9%
-17,886
-21.5%
10
3.0%
ICT manufacturing total ICT SERVICE INDUSTRIES Wholesale of ICT products
5165 : Wholesale: other industry machinery etc
133,307
10.6%
132,761
1 1 .5%
1,649
8.4%
1,688
9.7%
-546
-0.4%
39
2.4%
Total wholesale
260,848
20.7%
238,015
20.6%
2,780
14.1%
2,847
16.4%
-22,833
-8.8%
67
2.4%
6420 : Telecommunications
244,053
19.4%
241,068
20.8%
3,786
19.3%
3,802
21.9%
-2,985
-1.2%
16
0.4%
2,901
0.2%
3,677
0.3%
nd
nd
nd
nd
776
26.7%
nd
nd
16,892
1.3%
19,283
1.7%
nd
nd
64
0.4%
2,391
14.2%
nd
nd
7133 : Renting of office machinery and equip. Computer software and services 7210 : Hardware consultancy 7220 : Software consultancy and supply
287,428
22.8%
299,885
25.9%
4,577
23.3%
4,571
26.3%
12,457
4.3%
-6
-0.1%
7230 : Data processing
55,254
4.4%
37,334
3.2%
nd
nd
nd
nd
-17,920
-32.4%
nd
nd
7240 : Data base activities
10,535
0.8%
10,879
0.9%
nd
nd
nd
nd
344
3.3%
nd
nd
7250 : Maintenance/repair: office machinery etc 7260 : Other computer related activities Total computer services
19,095
1.5%
26,898
2.3%
nd
nd
nd
nd
7,803
40.9%
nd
nd
112,544
8.9%
91,898
7.9%
661
3.4%
572
3.3%
-20,646
-18.3%
-89
-13.5%
501,748
39.8%
486,177
42.0%
5,777
29.4%
5,247
30.2%
-15,571
-3.1%
-530
-9.2%
ICT services total
1,009,550
80.2%
968,937
83.8%
12,343
62.8%
11,896
68.4%
-4.0%
-447
-3.6%
ICT SECTOR: TOTAL
1,259,532
100.0%
1,156,746
100.0%
19,661
100.0%
17,390
100.0%
-40,613 102,786
-8.2%
-2,271
-11.6%
Note: nd = not disclosed due to confidentiality constraints; For this reason NI ICT sector totals do not include headings 3130, 3330 and 7133. Sources: GB data from ONS Annual Business Inquiry (via NOMIS). NI data from 2001 and 2003 Census of Employment, supplied by DETI Statistics Research Branch.
17
3. The Spatial Distribution of Division 72 Employment at Regional, County and Local Scales
3.1 Overview a. Having
reviewed
national
trends
in
Division
72
employment,
and
contextualised them within the wider ICT sector, this chapter turns to exa mine the spatial (i.e. geographical) distribution of “computer software and services” jobs across the UK. This spatial analysis is conducted across three different spatial scales: regional, county and local. b. First, the regional pattern is analysed for the nine English Government Office Regions (London, South East, East of England, South West, West Midlands, East Midlands, Yorkshire and the Humber, North West and North East) plus Scotland and Wales. Northern Ireland and Republic of Ireland, which have separate data series, are then introduced to the comparison.
c. Second, the sub-regional pattern of Division 72 employment is examined using the 46 English and 8 Welsh counties, and 12 Scottish regions, which existed in GB prior to the introduction of unitary authorities in 1996. These 66 “counties” – which are depicted on the map in Appendix B - give an intermediate scale of analysis between the regional and the local. In the absence of county-level data for Northern Ireland, two NI sub-regions (“Greater Belfast” and the “Rest of NI”) are introduced into the analysis. d. Third, the local pattern of Division 72 employment is examined at the scale of local government areas. Since the 1996 local government re-organisation, GB has been divided into 408 Unitary Authorities and Local Authority Districts (UALADs), which are shown by the maps in Appendix Ci and Cii. Data for NI’s five NUTS III regions are introduced into the local analysis in the absence of district council level data (Appendix Ciii).
3.2 The Regional Scale
18
3.2.1 The GB Regions e. At the regional scale of analysis, the spatial distribution of Division 72 employee jobs is highly uneven (Table 3.1). In 2003, London and the South East – which had 97,600 and 123,900 employees respectively – together accounted for just under half of all GB computer software and services employees. And two further regions – East and North West, with around 50,000 employees each – accounted for a further fifth. The remaining seven GB regions accounted for only one third of GB computer software and services employees.
f.
The dominance of the ‘Greater South East’ (GSE - i.e. London, South East and East regions) has been a hallmark of the GB computer software and services industry since its foundation in the 1960s (Coe, 1996a). Official employment data show the persistence of this dominance for more than a decade. The GSE consistently accounted for around 60% of GB Division 72 employees until 2001 after which it fell slightly 8 .
g. Overall, the regional Division 72 employee hierarchy displayed remarkable stability over the period 1995-2003 (Table 3.1 and Figures 3.1 and 3.2). Despite the fact that GB Division 72 employment more than doubled over the period, all the regions bar Scotland (up 2 places) and East Midlands (down 2) were ranked in the same place in 2003 as they had been in 1995.
h. All the GB regions experienced strong net employment growth in Division 72 during the period 1995-2003. However, the average growth rate was lowest in London (CAGR 7.5%) and East region (CAGR 7.6%) and highest in the North East (CAGR 14.8% and Scotland (CAGR 13.8%). North West (12%) and Yorkshire and the Humber (11.6%) also had growth rates significantly above the GB average (Table 3.1 and Figures 3.1 and 3.2). Five regions (South East, South West, West Midlands, East Midlands and Wales) had growth rates close to (+/-1%) the overall GB CAGR of 9.5%. Later, Section 5 explores this regional pattern of employment change in more detail.
8
The GSE share of GB Division 72 was 59.1% in 1991, 60.7% in 1995 and 59.3% in 2001 but fell to 56.0% in 2003. As Section 5 later shows, this change reflects the differential regional impact of the IT downturn. It remains to be seen whether this will be a permanent shift or merely a ‘blip’.
19
Table 3.1: Ranking of Government Office Regions by Division 72 Employees in 1995 and 2003 1995 Ranks
Region
1995 employees
% of UK
2003 Ranks
Region
2003 employees
% of UK
CAGR 1995-03
change in rank
1
South East
60,493
25.4%
1
South East
123,869
25.2%
9.4%
0
2
London
3
Eastern
54,913
23.1%
2
28,072
11.8%
3
London
97,596
19.9%
7.5%
0
Eastern
50,583
10.3%
7.6%
4
North West
19,605
8.2%
0
4
North West
48,603
9.9%
12.0%
0
5
West Midlands
17,148
6
South West
17,074
7.2%
5
West Midlands
36,760
7.5%
10.0%
0
7.2%
6
South West
32,938
6.7%
8.6%
7
East Midlands
0
11,884
3.9%
7
Scotland
26,389
5.4%
13.8%
+2
8
Yorkshire and The Humber
10,417
4.4%
8
Yorkshire and The Humber
25,133
5.1%
11.6%
0
9 10
Scotland
9,354
5.0%
9
East Midlands
24,636
5.0%
9.5%
-2
North East
3,864
1.6%
10
North East
11,681
2.4%
14.8%
11
0
Wales
3,635
1.5%
11
Wales
7,987
1.6%
10.3%
0
236,459
99.4%
486,175
98.9%
9.4%
1,380
0.6%
5,247
1.1%
18.2%
237,839
100.0%
491,422
100.0%
9.5%
London & the South East
115,406
48.5%
London & the South East
221,465
45.1%
8.5%
GSE (i.e. SE, LO, EA)
GB total 12
Northern Ireland UK total
GB total 12
Northern Ireland UK total
Selected aggregates
0
Selected aggregates 143,478
60.3%
GSE (i.e. SE, LO, EA)
272,048
55.4%
8.3%
English Midlands (WM, EM)
29,032
12.2%
English Midlands (WM, EM)
61,396
12.5%
9.8%
North England (NW, YH, NE)
33,886
14.2%
North England (NW, YH, NE)
85,417
17.4%
12.3%
Celtic Fringe (SC, WA, NI)
14,369
6.0%
Celtic Fringe (SC, WA, NI)
39,623
8.1%
13.5%
(2002 data)
Republic of Ireland
5,381
Republic of Ireland Southern & Eastern region Border Midlands West region
19,678
% RoI
18,130
92.1%
1,548
7.9%
20.4%
(1995-02)
Note: Republic of Ireland data are “number of employees” and are used for comparability with ABI “employees in employment” data. The ASI also gives figures for “number of persons engaged” (i.e. including proprietors and family members) as follows: 5,823 (1995) and 22,210 (2002). Source: author’s calculations using data from ONS Annual Business Inquiry, NI Quarterly Employment Survey and CSO Annual Services Inquiry.
20
Figure 3.1: Division 72 Employees by Region, 1995-2003 140,000 gor:South East
gor:London
120,000
gor:Eastern
100,000
gor:North West
gor:West Midlands
80,000 gor:South West
gor:Scotland
60,000 gor:Yorkshire and The Humber
gor:East Midlands
40,000
gor:North East
20,000
gor:Wales
Northern Ireland
0 1995
1996
1997
1998
1999
2000
2001
2002
2003
Source: ONS Annual Business Inquiry and NI Census of Employment/Quarterly Employment Survey.
Figure 3.2: Division 72 Employees by Region (excluding GSE), 1995-2003 50,000 gor:North West
gor:West Midlands
40,000 gor:South West
gor:Scotland
30,000
gor:Yorkshire and The Humber
20,000
gor:East Midlands
gor:North East
10,000 gor:Wales
Northern Ireland
0 1995
1996
1997
1998
1999
2000
2001
2002
2003
Source: ONS Annual Business Inquiry and NI Census of Employment/Quarterly Employment Survey.
21
3.2.2 Northern Ireland i.
The trajectory of NI Division 72 employment during the period September 1995 to March 2004, as reported by the NI Quarterly Employment Survey, is depicted in Figure 3.3. The number of computer software and services employees increased rapidly from only 1,380 in September 1995 to a peak of 5,780 in September 2001. As with the GB trend, Division 72 employment in NI is then said to have declined to around 5,100 in June 2003 (a fall of 11.4% from the peak) before recovering slightly to 5,360 in March 2004 (and 5,750 in September 2004) 9 . Thus, NI had fewer Division 72 employees than any of the GB regions throughout the period. This is not unexpected since NI is the smallest UK region. The location quotient analysis in Chapter 4 allows us to say more about the relative size of Northern Ireland’s computer software and services sector in the UK regional context .
j.
As noted in Section 1, recent research by ERINI’s Priority Skills Unit (PSU) has raised doubts over the accuracy of official employment estimates for the sector. Specifically, on the basis of an extensive telephone survey (covering most of the key players), McGuinness and Doyle (2003, p.7) estimated NI Division 72 employment at 6,980 in November 2002, which is over 30% higher than the December 2002 QES figure of 5,260 (Figure 3.3).
k. What we can say, regardless of which figures are used, is that NI was clearly the fastest growing UK region in terms of Division 72 employees between 1995 and 2003 (CAGR 18.2% according to QES data, obviously higher on the basis of the ERINI PSU figures). One interpretation is that NI was belatedly beginning to ‘catch-up’ with other GB regions from a low starting point. However, the fact that NI was able to outperform the two regions immediately above it in the regional league table - particularly Wales – should not be downplayed.
9
Source: September 2004 QES 2 digit analysis from DETI website (accessed July 2005).
22
Figure 3.3: NI Division 72 Employment, September 1995-March 2004 8,000 ERINI PSU est. 7,000
NI CoE 2003
6,000
5,000
4,000 ERINI PSU est. 3,000
2,000
1,000
0 Sep-95
Sep-96
Sep-97
Sep-98
Sep-99
Sep-00
Sep-01
Sep-02
Sep-03
Note: The chart shows QES data as supplied by DETI Statistics Research Branch in July 2004. These data were supplied with the warning that data from December 2001 onwards were subject to revision following the release of the 2003 NI CoE. The subsequently released NI CoE 2003 figure is shown on the chart for reference purposes. Any related revisions to the QES had not been obtained at time of writing. Sources: NI Quarterly Employment Survey, 1995-2004; NI Census of Employment 2003; ERINI Priority Skills Unit (McGuinness and Bonner, 2000; McGuinness and Doyle, 2003).
3.2.3 The Republic of Ireland
l.
The rapid expansion of the Republic of Ireland’s software industry during the 1990s attracted the attention of policy-makers in competitor regions and has now been extensively studied by academics (e.g. Coe, 1997 and 1999; O’Malley et al, 1997; Ó Riain, 1999; Crone, 2002a and 2002b). For this reason, and due to its geographical proximity and land border with NI, the Republic (hereafter ROI) is a logical benchmark for the regions and subregions of the UK, and for NI in particular.
m. According to data from the CSO’s Annual Services Inquiry, Division 72 employment in ROI grew from only 5,380 in 1995 to over 19,680 in 2002 (Table 3.1) 1 0 . In 2002, Heading 223 “Reproduction of recorded media” 10
2003 data were not available at the time of writing. The ASI also gives figures for the “number of persons engaged” in Division 72: e.g. 5,823 in 1995; 22,210 in 2002. The difference can be attributed to the inclusion of business proprietors and family members in the “persons engaged” figures. There are apparently at least several hundred small family firms or sole traders in the ROI industry.
23
(which is dominated by Heading 2233 “Reproduction of computer media” covering the reproduction of packaged software) employed 6,000 further persons (up from 4,000 in 1995). Thus, the ROI computer software and services industry can be said to have employed at least 25,000 people in 2002 (Figure
3.4).
In
simple
employment
terms,
therefore, th
comparable to Scotland and Yorkshire and the Humber, the 7
ROI
is
and 8th
regions in the UK regional hierarchy. Certainly the ROI industry is small by the standards of South East England. However, these simple comparisons mask some of the interesting and important structural features of the ROI industry; notably its high export orientation and focus on software products (Crone, 2002a). n. Arguably the most notable feature of the ROI software industry, however, is its rapid growth since the mid-1990s. Division 72 employment in ROI grew at a CAGR of 20.4% over the period 1995-2002. Among the 12 UK regions, only Northern Ireland matched this high growth rate (21.7% over the same period).
Figure 3.4: Computer software and services employees in the Republic of Ireland, 1995-2002
NACE 223 Reproduction of recorded media
NACE 72 Computer services
30,000
25,000
20,000
15,000
10,000
5,000
0 1995
1996
1997
1998
1999
2000
2001
Source: Annual Services Inquiry and Census of Industrial Production, Central Statistics Office.
24
2002
o. Recent
editions
of
the
ASI
have
also
included
some
sub-regional
employment data for Division 72. Ireland is now divided into two NUTS2 regions: the Southern and Eastern region (covering the more affluent parts of the country including the capital Dublin and leading provincial cities Cork and Limerick) and the Border, Midlands and West region. The ASI shows that, in 2002, ROI Division 72 employment was highly concentrated in the S&E region (92%) with only 1,550 employees in the BMW region (Table 3.1) 1 1 . Research by Crone (2002a) and others has shown that the majority of employees (and the leading firms) are to be found in Greater Dublin.
3.3 The County Scale p. The regional scale of analysis gives a broad, if somewhat crude, impression of the distribution of employment but it masks considerable spatial variations within each region. The county scale of analysis further exposes the highly uneven geographical distribution of computer software and services jobs within the UK (Table 3.2) 1 2 . For example, London (treated here as a county) and the next seven counties at the top of the ranked county employment distribution (or hierarchy) together accounted for more than half of all UK Division 72 employees in 2003. Further, if the county hierarchy is divided into three equal groups, we find that the top third covered 78.4% of UK Division 72 employees in 2003, the middle third covered another 17.9%, and the bottom third covered only 3.6%.
q. The high level of spatial concentration of Division 72 employees observed in 2003 is a persistent phenomenon. The figures for the percentage of employees located in London and the next 7 counties, and for the percentage in the top, middle and bottom third of the county hierarchy, are almost unchanged since 1995 (Table 3.2).
r.
Apart from London - which had 97,600 Division 72 employees (20% of the UK total) in 2003 – the Top 10 counties in 2003 comprised five ‘nonmetropolitan’ counties located in a contiguous ‘arc’ or ‘crescent’ to the
11
For comparison note that the S&E region had 73.5% of the ROI population in the 2002 Census and 73.6% of all ROI manufacturing employees in the 2001 Census of Industrial Production. 12
Recall that GB is divided into a total of 66 ‘counties’ (46 English and 8 Welsh counties plus 12 Scottish regions) for the purpose of this analysis, and NI into two sub-regions (Greater Belfast and Rest of NI).
25
western side of London (Berkshire, Surrey, Hampshire, Hertfordshire and Buckinghamshire), three ‘metropolitan’ counties located further north (West Midlands, Greater Manchester and West Yorkshire) plus Avon (Table 3.2). Berkshire, the leading GB County apart from London, had around 35,000 computer software and services employees in 2003 (7. 1% of the UK total and more than NI and ROI combined!). Greater Manchester, the leading northern
county,
had
over
21,000
computer
software
and
services
employees in 2003. The nine leading counties in 2003 were identical to the leading nine counties from 1995, illustrating the persistence of the spatial hierarchy in UK computer software and services (Table 3.2).
s. Employment growth rates at the county scale showed greater variation than was seen at the regional scale, as we might expect. Focusing on members of the 2003 Top 20 (per Table 3.2), the fastest growing counties over the period 1995-2003 were Lancashire (CAGR 19.4%, up 16 places in the hierarchy), Lothian (CAGR 16.1%, up 9 places) and Strathclyde (CAGR 13.3%, up 2
places). In contrast, the slowest growing county was
Hertfordshire (CAGR 6. 1%, down 1 places). t.
In the absence of county-level data, NI can be divided into two sub-regions for a cross-sectional comparison with the GB counties (Tables 3.2 and 3.4). According to the 2003 NI Census of Employment “Greater Belfast” (an aggregation of the Belfast and Outer Belfast NUTS III regions – i.e. Belfast, Castlereagh, Newtownabbey, Carrickfergus, North Down and Lisburn district councils) had 3,923 Division 72 employees in September 2003, which ranked it 35th out of 68 UK counties. For comparison, note that the metropolitan counties of Merseyside, Tyne & Wear and South Yorkshire had between 5,000 and 6,500 computer software and services employees in 2003. Second, the Rest of NI sub-region had only 1,324 Division 72 employees in 2003 (ranked 51s t of 68 UK counties), which was only slightly more
than
the
English
counties
of
Lincolnshire
and
Cornwall
considerably fewer than the likes of Cumbria, Norfolk and Humberside.
26
and
Table 3.2: Ranking of Pre-1996 Counties/Scottish regions by Division 72 Employees in 1995, 2001 and 2003 rank
pre-1996 county / scottish region
2003 empl.
% of UK
1
Greater London
54,913
Greater London
2
Berkshire
14,726
Berkshire
36,770
Greater London
97,596
19.9%
42,683
7.5%
0
Berkshire
34,934
7.1%
20,208
11.4%
3
Surrey
13,305
Surrey
0
27,050
Surrey
24,790
5.0%
11,485
8.1%
0
4
Hampshire
11,024
5
Hertfordshire
10,889
Hampshire
21,844
Hampshire
23,295
4.7%
12,271
9.8%
0
Hertfordshire
18,761
Greater Manchester
21,413
4.4%
11,925
10.7%
+1
6
Greater Manchester
9,488
Greater Manchester
17,896
Hertfordshire
17,492
3.6%
6,603
6.1%
-1
7 8
West Midlands
9,044
West Midlands
17,194
West Midlands
17,203
3.5%
8,159
8.4%
0
Buckinghamshire
8,306
Buckinghamshire
14,593
Buckinghamshire
15,795
3.2%
7,489
8.4%
0
9
West Yorkshire
5,630
Strathclyde region
11,876
West Yorkshire
12,830
2.6%
7,200
10.8%
0
10
Essex
5,604
Avon
10,972
Avon
11,927
2.4%
6,363
10.0%
+1
11
Avon
5,564
Cambridgeshire
10,817
Essex
11,338
2.3%
5,734
9.2%
-1
12
Cambridgeshire
4,831
West Yorkshire
10,765
Cambridgeshire
10,556
2.1%
5,725
10.3%
0
13
Cheshire
4,724
Lancashire
9,808
Lancashire
9,826
2.0%
7,454
19.4%
+16
14
Nottinghamshire
3,857
West Sussex
9,162
Cheshire
9,609
2.0%
4,885
9.3%
-1
15
Wiltshire
3,829
Essex
8,776
Strathclyde region
9,579
1.9%
6,045
13.3%
+2
16
Bedfordshire
3,676
Lothian region
8,542
Nottinghamshire
9,253
1.9%
5,396
11.6%
-2
17
Strathclyde region
3,534
Oxfordshire
8,228
Lothian region
8,569
1.7%
5,966
16.1%
+9
18
Oxfordshire
3,466
Cheshire
8,164
Oxfordshire
7,578
1.5%
4,112
10.3%
0
19
West Sussex
3,447
Wiltshire
7,788
Kent
6,834
1.4%
3,435
9.1%
+1
20
Kent
3,399
Nottinghamshire
7,567
West Sussex
6,612
1.3%
3,165
8.5%
-1
Greater Belfast (rk41)
1,030
Greater Belfast (rk35)
4,105
Greater Belfast (rk35)
3,923
0.8%
2,893
+6
Rest of NI (rk47)
1,672
Rest of NI (rk51)
1,324
0.3%
974
18.2%
Rest of NI (rk57)
1995 empl.
350
pre-1996 county / scottish region
2001 empl. 119,814
pre-1996 county / scottish region
Top 10 (ex. London)
39.3%
Top 10 (ex. London)
37.0%
Top 10 (ex. London)
191,017
38.9%
London + next 7
55.4%
London + next 7
54.0%
London + next 7
252,518
51.4%
Top 23
80.6%
Top 23
80.1%
Top 23
385,460
78.4%
Next 23
16.6%
Next 23
16.9%
Next 23
88,154
17.9%
Bottom 22
17,807
3.6%
Bottom 22
2.8%
Bottom 22
3.1%
Changes 1995-2003 empl. CAGR rank
+6
Note: Greater Belfast=Belfast+Outer Belfast NUTSIII regions; Rest of NI=East of NI+North of NI+West and South of NI NUTS III regions; 1995 data are author’s estimates. Source: author’s calculations using data from ONS Annual Employment Survey 1995, Annual Business Inquiry 2001 & 2003, NI Census of Employment 2001 & 2003.
27
3.4 The Local Scale u. The local scale gives a further, more detailed insight into the distribution of computer software and services employment across the UK. The high level of spatial concentration is again the main feature when we look at the 2003 UALAD hierarchy (Table 3.3). The Top 20 UALADs together accounted for 26.6% of all GB Division 72 employees in 2003, and the top 50 for 47.6%. In contrast, the bottom 50 UALADs in the hierarchy accounted for just 1.2% of
UK
Division
72
employees
in
2003.
These
measures
of
spatial
concentration are almost unchanged from 1995.
v. Among the Top 20 UALADs in 2003, in terms of Division 72 employment, were 6 London Boroughs and 6 Unitary Authorities from the ‘Western Arc’ of the South East region (Table 3.3). The remaining 8 UALADs were provincial cities outside the GSE: Leeds in Yorkshire, Birmingham in West Midlands, Edinburgh
and
Glasgow
in
Scotland,
Nottingham
in
East
Midlands,
Manchester and Trafford in the North West and Bristol in the South West. Note that only four UK regions had no representation in the 2003 Top 20: East (South Cambridgeshire ranked 32nd), North East (Newcastle 38th), Northern Ireland (Belfast 44th) and Wales (Cardiff 100th and Swansea 113th). The spatial distribution of computer software and services employment at UALAD level is also illustrated in Figures 3.5-3.7, which map Division 72 employee density (i.e. the number of employees per square kilometre) in 2001 – the peak year for UK employment in the sector.
w. At the top end of the UALAD distribution, employment growth rates were of a similar order of magnitude to those at the county scale. Among those provincial cities in 2003 Top 20, Glasgow, Edinburgh and Nottingham experienced the fastest growth in Division 72 employment over the period 1995-2003 – each moved up more than 10 places in the UALAD ranking (Table 3.3). The London boroughs of Islington and Hillingdon were among the worst performing UALADs among the 2003 Top 20. Note also that, at the bottom end of the distribution, 15 UALADs actually experie nced a net loss of Division 72 employees during 1995-2003, and a further 65 experienced CAGR of less than half the UK rate (i.e. 100 indicates an area has more Division 72 employees than might be expected on the basis of its share of national employment in all industries (i.e. relative specialisation);
§
LQ=100 indicates an area has a share of Division 72 employment in accordance with its share of national employment in all industries;
§
LQ100) with the South East clearly leading London and East. These LQs confirm the GSE’s domination of UK computer software and services employment in both absolute and relative terms. The second tier in 2003 (LQ 80-90) comprises three regions stretching down the western side of England: North West, West Midlands and South West (Figures 4.2 and 4.3). The third tier (LQ 6075) comprises four regions in the northeastern part of the UK: East Midlands, Yorkshire and the Humber, North East and Scotland. The final tier (LQ170 (shaded purple), we can de-limit a swathe of geographically contiguous localities with a high degree of specialisation in computer software and services (see Appendix E for details
and
extra
maps).
This
crescent-shaped
area
extends
from
Cambridge and South Cambridgeshire at its northern tip to Horsham in the south; and from West Berkshire in the west to west London in the east. In 2003, this area comprised 37 contiguous UALADs with one quarter of all UK Division 72 employees (122,000). Further, we can also identify a smaller geographical area - the Inner Core of the Western Arc – comprising 17 UALADs with LQ>250. This core area – which falls within a 40 mile radius of Bracknell and Wokingham, the UALADs with the greatest specialisation in Division 72 (LQs>600) – accounted for 14% of UK computer software and services employees in 2003 (66,800); more than twice as many employees as there were on the whole island of Ireland at this time.
o. Understanding the pre-eminence of the Western Arc is a complex question and largely beyond the scope of this report. Coe (1996a) points to the long-established
roots
of
the
industry
there,
dating
back
to
the
decentralisation of the early computer service firms from London in the 1960s, and subsequent processes of cumulative causation relating to high rates of new firm formation allied to locational inertia, plus the location decisions
of
leading
multinationals
companies
like
Microsoft,
Oracle
processes
are
undoubtedly
(e.g. and
reinforced
European
Computer by
some
headquarters
Associates). specific
of
These
location
advantages. Most notably, the Western Arc has excellent transport links (e.g. M4, M3, M25 and M40 motorways; proximity to Heathrow airport) which permit easy access to domestic and international customers. The GSE is also a focal point of demand for computer software and services, being the dominant UK location for corporate HQs and the important financial services sector. Firms located in the Western Arc might also benefit from “localised agglomeration economies” which can be derived where there is a “critical mass” of similar or related firms (e.g. access to a large pool of skilled/experienced labour, localised knowledge spillovers).
42
Table 4.3: Rankings of GB UALADs and NI NUTS III regions by Division 72 Location Quotients in 1995 and 2003 Rank
UALAD (region)
Rank
UALAD (region)
1
Bracknell Forest (SE)
LQ’95 853
1
Wokingham (SE)
758
2
Hart (SE)
825
2
Hart (SE)
742
3
Surrey Heath (SE)
688
3
Bracknell Forest (SE)
610
4
Wokingham (SE)
636
4
Runnymede (SE)
526
5
Rushmoor (SE)
497
5
Windsor & Maidenhead (SE)
450
6
Wycombe (SE)
467
6
Surrey Heath (SE)
380
7
Spelthorne (SE)
415
7
Stevenage (EA)
346
8
Runnymede (SE)
395
8
Woking (SE)
346
9
Elmbridge (SE)
370
9
South Cambridgeshire (EA)
331
10
Three Rivers (EA)
370
10
Spelthorne (SE)
321
11
Woking (SE)
362
11
West Berkshire (SE)
319
12
Dacorum (SE)
360
12
Rushmoor (SE)
319
13
Windsor & Maidenhead (SE)
336
13
Wycombe (SE)
297
14
South Cambridgeshire (EA)
330
14
Dacorum (SE)
294
15
Winchester (SE)
322
15
Fylde (NW)
290
16
West Berkshire (SE)
313
16
Warwick (WM)
285
17
Redditch (WM)
309
17
South Buckinghamshire (SE)
277
18
Slough (SE)
294
18
St Albans (EA)
269
19
Welwyn Hatfield (EA)
285
19
Mole Valley (SE)
268
20
Islington (LO)
282
20
Milton Keynes (SE)
266
21
Richmond-upon-Thames (LO)
278
21
Waverley (SE)
256
22
South Oxfordshire (SE)
270
22
Slough (SE)
253
23
Hounslow (LO)
268
23
Elmbridge (SE)
253
24
Kingston-upon-Thames (LO)
265
24
Trafford (NW)
252
25
Hillingdon (LO)
259
25
Hounslow (LO)
244
26
North Hertfordshire (EA)
259
26
Basingstoke and Deane (SE)
234
27
Chiltern (SE)
259
27
Portsmouth (SE)
234
28
East Hampshire (SE)
257
28
Kingston-upon-Thames (LO)
222
29
St Albans (EA)
248
29
Solihull (WM)
221
30
Merton (LO)
247
30
South Oxfordshire (SE)
221
Next 10 highest northern UALADs
LQ’03
Next 10 highest northern UALADs
45
Trafford (NW)
192
32
Telford and Wrekin (WM)
205
46
Congleton (NW)
190
44
Nottingham (EM)
178
49
Warrington (NW)
187
50
Stockport (NW)
167
53
Stratford-on-Avon (WM)
179
54
Chorley (NW)
161
54
Erewash (EM)
177
55
Rushcliffe (EM)
160
68
Warwick (WM)
166
56
Macclesfield (NW)
159
74
North East Derbyshire (EM)
160
71
Congleton (NW)
142
84
Rushcliffe (EM)
144
72
Lichfield (WM)
141
90
Bromsgrove (WM)
142
74
Durham (NE)
139
91
Vale Royal (NW)
138
75
Bromsgrove (WM)
139
Northern Ireland NUTS III regions 156
Belfast
82
260
North of NI
49
271
Outer Belfast
46
407
East of NI
13
413
West & South of NI
Notes: no data available for NI NUTS III regions for 1995. Northern UALADs inside top 30 in bold. Source: author’s analysis of data from ONS Annual Business Inquiry and NI Census of Employment.
43
7
Figure 4.9: Div ision 72 UALAD LQ Map for Southern Britain, 2001
Source: author’s analysis of Annual Business Inquiry.
Figure 4.10: Division 72 UALAD LQ Map for Southern Britain, 2003
Source: author’s analysis of Annual Business Inquiry.
44
p. The UALAD LQ analysis also highlights the highly uneven pattern of Division 72 employment within southeastern England. Thus, in contrast to the Western Arc, the eastern counties of the GSE contain only a handful of UALADs
with
a
2003
Division
72
LQ>100.
(e.g.
Basildon,
LQ
192;
Chelmsford, 155; Sevenoaks, 133; Brentwood, 130; Bromley, 126). In fact, there are many UALADs in this part of England with LQ100) in 2003 were clusters of UALADs around the southern fringe of Manchester in the North West (notably Trafford, LQ 252; Stockport, 168; Macclesfield, 158; Congleton, 142; Halton, 136; and Warrington, 131) and around the southern and eastern fringe of Birmingham in the West Midlands (notably Warwick, LQ 285; Solihull, 221; Lichfield, 141; Bromsgrove, 139; and Stratford-upon-Avon, 136). Other Northern hotspots in 2003 included the city of Nottingham and nearby Rushcliffe and Broxtowe in the East Midlands (LQs 178, 159 and 130), City of Edinburgh and West Lothian in Scotland (LQs 107 and 133), Harrogate and Leeds in Yorkshire (LQs 107 and 101) and Newcastle-upon-Tyne (106). Among the more unusual hotspots were Fylde and Chorley in Lancashire and Telford & Wrekin in Shropshire and Durham in the North East (all LQs>130).
45
Figure 4.11: Division 72 UALAD LQ Map for Northern Britain, 2001
Source: author’s analysis of Annual Business Inquiry and NI Census of Employment.
Figure 4.12: Division 72 UALAD LQ Map for Northern Britain, 2003
Source: author’s analysis of Annual Business Inquiry and NI Census of Employment.
46
s. Inclusion of the five NI NUTSIII regions in the UK UALAD LQ analysis allows us to comment on their relative specialisation in computer software and services in 2001 and 2003 (see Figures 4.11 and 4.12 and Table 4.3). Belfast,
the
dominant
location
for
computer
software
and
services
employment within NI, had a 2003 LQ of 82, and was ranked 156th out of 413 UK UALADs. Belfast was placed just above Sheffield in the middle of 18 provincial cities (as used previously in Section 3), in terms of its relative specialisation in computer software and services (Figure 4.13). Thus, Belfast was much better placed in the sector, in both absolute (as per Section 3) and relative terms, than the cities of Aberdeen, Swansea, Liverpool, Hull, Bradford, Cardiff and Leicester. However, Belfast trailed well behind the cities of Nottingham, Bristol, Edinburgh, Newcastle-upon-Tyne, Leeds and Manchester in both absolute and relative terms.
t.
The other four NI NUTSIII regions compared much less favourably than Belfast, having 2003 LQs ranging from 49 in the North of NI - ranked 260th out of 413 UK UALADs - to only 7 in the West & South of NI – the lowest of all the 413 UK UALADs (Table 4.3).
Figure 4.13: Division 72 LQs for Belfast and 18 GB Cities, 2001 and 2003 200
LQ 2001 LQ 2003 150
100
Source: author’s analysis of Annual Business Inquiry and NI Census of Employment.
47
Leicester City
Cardiff
Bradford
Kingston upon Hull
Liverpool
Swansea
Glasgow City
Birmingham
Aberdeen City
Sheffield
Belfast
Derby City
Coventry
Manchester
Leeds
Edinburgh, City of
Bristol
Nottingham
0
Newcastle-upon-Tyne
50
5. Recent Changes in Division 72 Employment: Investigating the Spatial Impact of the IT Downturn
5.1 Overview a. Recall from Section 2.1 that, nationally, Division 72 employment underwent a period of sustained growth between 1995 and 2001, with particularly high growth rates (over 15% per annum) between 1995 and 1999. Growth then slowed slightly in 2000 and fell to only 2.5% during 2001. The year 2002 was a watershed year for the sector with total UK employment falling by around 3.4% from the 2001 peak of 507,500. The decline in employment was short-lived, however, as the UK as a whole experienced a marginal increase in Division 72 employment during 2003
b. The aim of this section is to examine the geographical dimensions of these national
employment
trends,
with
a
particular
focus
on
the
spatial
employment impact of the downturn (or slowdown) in the global ICT industries. As the downturn was first detected during the year 2000 (e.g. stock market declines in the telecoms and technology sectors) we are in a position to examine its impact over a three year period from December 2000 to December 2003. The analysis below follows the precedent of Chapters 3 and 4 by examining recent changes in Division 72 employment at the regional, county and local scales.
5.2 Recent Changes at the Regional Scale c. Nationally, Division 72 employment grew every year from 1995 to 2001 before declining during 2001-02 (Figure 2.1). But how did these trends manifest themselves at the regional scale? The majority (nine) of the UK regions
experienced
continuous
year-on-year
growth
in
Division
72
employment throughout the period 1995 to 2000 (Figure 5.1 and maps in Figures 5.3-5.7) 1 6 . Focusing on the three years after December 2000 16
The only UK regions where this was not the case were Wales, in 1995-96 and 1999-2000; Scotland, in 1996-97; and the No rth East, in 1998-99.
48
(Figures 5.2, 5.8-5.10) we find that the downturn had a regionally differentiated impact on computer software and services employment. More specifically, Division 72 employment: ⇒ declined in London in all three years (2000-01, 2001-02 and 2002-03); ⇒ declined in 2001-02 and 2002-03 in South East, Scotland and Northern Ireland (with the more severe decline in 2001-02 in all cases); ⇒ declined in 2001-02 but returned to growth in 2002-03 in South West, East Midlands and Wales; ⇒ declined in 2000-01 but returned to growth thereafter in East; ⇒ continued to expand in all three years (but at a slower rate than previously) in the North West and Yorkshire and the Humber; ⇒ and continued to expand in 2000-01 and 2001-02, with only a marginal decline in 2002-03, in West Midlands. As we only have data for 2000-01 and 2001-02, we cannot fully explore the impact of the downturn on the Republic of Ireland. However, the official data do not show any decline in Division 72 employment during 2001-02 (Figure 5.2).
d. In absolute terms, note that London and the South East regions bore the brunt of the downturn. Division 72 employment in London declined by 26,800 during 2000-03 (a fall of 22%). The South East also lost 6,000 Division 72 jobs during 2001-03 (a fall of 5%). Northern Ireland and Scotland were also relatively harshly hit during the downturn period, losing 9.2%
and
4.1%
of
their
Division
72
employment
during
2001-03
respectively. In contrast, the North of England appears to have emerged from the downturn period unscathed, as Division 72 employment actually continued to rise during the period. The reasons for this regionally differentiated pattern are a matter of conjecture at this time.
One
hypothesis is that the industry in the GSE is more exposed to global forces than the English North due to the presence of many leading multinationals in the former. GSE-based computer software and service firms may also be more dependent on sales to sectors most affected by the global slowdown (e.g. financial services, the telecoms and technology sector). Research by Coe (1996a) in the early 1990s found that Northern computer service firms were more oriented to manufacturing and public sector markets than their southern counterparts. Certainly the public sector market segment has been quite resilient during the IT downturn.
49
Figure 5.1: Year-on-year Percentage Change in Division 72 Employees by Region, 1995-2003 60% 1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
50%
40%
30%
20%
10%
0%
-10%
-20% South East
London
Eastern
North West
West Midlands
South West
East Yorkshire Midlands and The Humber
Scotland
Republic North East of Ireland
Wales
Northern Ireland
Source: analysis of ONS Annual Business Inquiry, NI QES and CoE, and CSO Annual Service Inquiry.
Figure 5.2: Year-on-year Percentage Change in Division 72 Employees by Region, 2000-2003 only 25%
2000-01
20%
2001-02
2002-03
15%
10%
5%
0%
-5%
-10%
-15% South East
London
Eastern
North West
West Midlands
South West
East Midlands
Yorkshire and The Humber
Scotland
Republic North East of Ireland
Wales
Northern Ireland
Source: analysis of ONS Annual Business Inquiry, NI QES and CoE, and CSO Annual Service Inquiry.
50
Figure 5.3: Percentage Change in Division 72 Employees by Region, 1995-1996
Figure 5.4: Percentage Change in Division 72 Employees by Region, 1996-1997
Figure 5.5: Percentage Change in Division 72 Employees by Region, 1997-1998
Figure 5.6: Percentage Change in Division 72 Employees by Region, 1998-1999
Figure 5.7: Percentage Change in Division 72 Employees by Region, 1999-2000
Figure 5.8: Percentage Change in Division 72 Employees by Region, 2000-2001
Figure 5.9: Percentage Change in Division 72 Employees by Region, 2001-2002
Figure 5.10: Percentage Change in Division 72 Employees by Region, 2002-2003
51
e. The
impact
of
the
IT
downturn
on
Northern
Ireland’s
Division
72
employment is a matter of debate. NI Census of Employment data, as reported above, suggest a contraction in employment of 9.2% between September 2001 and September 2003 (-530 jobs). However, McGuinness and Doyle (2003) provide somewhat contradictory evidence, at least for 2001-02. They estimated NI Division 72 employment at almost 7,000 in November 2002, implying continued growth during 2000-02 (Figure 5.11). McGuinness and Doyle (2003) also estimated the breakdown of NI Division 72 employment by ownership and suggested that employment in the foreign-owned segment of the NI computer software and services sector declined by up to a quarter during 2000-2002 but suggest this dip was offset by sustained employment growth in the indigenous sector.
Figure 5.11: ERINI PSU Estimate of NI Division 72 Employment, 1992-2002
Total Indigenous Foreign
Notes: Foreign-owned employment in 2000 was estimated using survey data and information on IDB supported firms in Division 72. Indigenous employment in 2000 was estimated as official Division 72 employment from QES minus foreign-owned employment. 2002 estimates were derived from survey data (i.e. as reported by surveyed companies, with grossing-up for non-respondents). Source: McGuinness and Doyle (2003), Figure 3, p.8.
5.3 Recent Changes at the County Scale f.
During the mid-to-late 1990s, Division 72 employment was experiencing strong growth nationally. At this time most counties were also experiencing
52
strong growth but even at this time not every county was experiencing employment growth (Figures 5.12-5.16). During the recent period of global uncertainty in the IT industry, which began in 2000 and worsened during 2001 and 2002, an increasing number of counties were seen to experience lower employment growth rates and even employment decline (Figures 5.17-5.19). The onset of this trend can be seen most clearly in Figure 5.20. By 2001-02 half of the GB counties were in decline and few were experiencing high rates of growth. g. The overall spatial pattern of recent county level changes in Division 72 employment is not especially coherent. Certainly we can see that large swathes of Southern England, Wales and Scotland experienced decline during 2001-02, which is consistent with the regional trends noted above (Figure 5.9). Similarly, many of the counties in the East region were in decline a year earlier in 2000-01 (Figure 5.17) and half of the counties in the South East were still in decline in 2002-03. Conversely, most counties in the English north and midlands were less severely affected, or unaffected, during the downturn period (Figures 5.17-5.19). However, there is no obvious logic to the recent pattern of growth and decline at the county scale.
h. Some interesting patterns are evident, though, if we looking at the fate of leading counties during 2000-01 and 2001-02, beginning in the ‘South’. Hertfordshire, Buckinghamshire and Surrey in the Western Arc all lost computer software and services jobs in both 2000-01 and 2001-02. As a result these three counties lost 16%, 15% and 5% of their Division 72 employees respectively during 2000-03. Berkshire lost jobs in 2001-02 and 2002-03 only, resulting in no net change in Division 72 employees during 2000-03. i.
The fate of the leading Northern metropolitan counties was rather more mixed. West Midlands and Merseyside lost jobs in both 2000-01 and 200203 (resulting in 2% and 7% declines respectively in Division 72 employment during 2000-03). In contrast, Division 72 employment grew by 26% and 15% respectively in Greater Manchester and West Yorkshire during 2000-03.
53
Figure 5.12: Percentage Change in Division 72 Employees by County, 1995-1996
Figure 5.13: Percentage Change in Division 72 Employees by County, 1996-1997
Figure 5.14: Percentage Change in Division 72 Employees by County, 1997-1998
Figure 5.15: Percentage Change in Division 72 Employees by County, 1998-1999
54
Figure 5.16: Percentage Change in Division 72 Employees by County, 1999-2000
Figure 5.17: Percentage Change in Division 72 Employees by County, 2000-2001
Figure 5.18: Percentage Change in Division 72 Employees by County, 2001-2002
Figure 5.19: Percentage Change in Division 72 Employees by County, 2002-2003
55
Figure 5.20: Number of Counties in Each Change Band, 1995-96 to 2002-03
1995-96
1
1996-97
1
5
4
1997-98
6
1999-00
6
2001-02
2002-03
4
17
6
7
1998-99
2000-01
9
8
3
>20% decrease
14
6
16
10
17
13
10-20% decrease
7
12
21
8
9
12
16
7
7
19
13
12
3
22
18
7
7
4
33
6
6
6
6
18
20
3
7
13
11
0-10% decrease
0-10% increase
6
5
16
10-20% increase
7
7
20-50% increase
3
6
>50% increase
Note: the chart summarises the information portrayed in Figures 5.3-5.10.
j.
The four Northern counties with the highest computer software and services LQs in 2003 (Cheshire, Nottinghamshire, Lothian and Warwickshire) fared very well during the three year downturn period. Division 72 employment grew by over 20% in each of these counties during 2000-03. Employment continued to grow in all three years in Cheshire and Warwickshire; Nottinghamshire
and
employment
2000-01
in
Lothian and
suffered
relatively
small
2001-02
respectively
but
declines these
in
were
outweighed by subsequent growth.
5.4 Recent Changes at the Local Scale k. With small geographical units such as GB UALADs there may be doubts about the reliability of year-on-year change data generated from a samplebased employment survey such as the ABI. Thus, any inferences on the impact of the downturn at this spatial scale should be viewed with caution. For this reason, this analysis looks only at recent changes in employment for the Western Arc as a whole (as identified in Section 4); and for Belfast and the leading provincial cities outside the Greater South East.
56
l.
First, recall that the Western Arc (as defined in Section 4.4.1) accounted for one quarter of all UK Division 72 employees (122,000) in 2003. Collectively, computer software and services employment in the 37 members of the Western Arc declined by 9.6% (nearly 13,000 jobs) in the period 2001-03. For comparison recall that total UK employment in Division 72 declined by 3.2% over the same period (~16,000 jobs). Thus, we can conclude that the Western Arc was disproportionately impacted by the IT downturn.
m. Second, as Figure 5.21 illustrates, the IT downturn had a highly variable impact on the leading provincial cities outside the GSE during 2001-03 (focusing on those cities with more than 3,000 employees in 2001). Thus, computer software and services employment continued to expand (to varying degrees) in Leeds, Sheffield, Nottingham, Manchester and Bristol but declined quite sharply (by more than 10%) in Birmingham, Edinburgh and Glasgow. Belfast had a more neutral experience with only a marginal (negative) change in its Division 72 employment between 2001 and 2003 1 7 .
Figure 5.21: Change in Division 72 Employment, 2001-2003, in Belfast and Leading Provincial Cities (with more than 3,000 Employees in 2001) 3,000
60%
change in employment, 2001-03 2,000
40%
Employment
1,000
20%
0
0%
-1,000
-20%
-2,000
-40%
-3,000
% change in employment
% change in employment, 2001-03
-60% Leeds
Nottingham
Sheffield
Bristol
Manchester
Belfast
Newcastle- Birmingham upon-Tyne
Edinburgh, Glasgow City City of
Source: author’s analysis of Annual Business Inquiry and NI Census of Employment.
17
Note, however, that the remainder of NI (i.e. outside Belfast) was much more adversely affected by the downturn, losing 17% of its Division 72 employees between 2001 and 2003 (see Table 3.4).
57
6. Analysis of Recent Foreign Direct Investment Patterns and Trends in the Software Sector
6.1 Overview a. This final substantive chapter looks at recent inward investment patterns and trends in the computer software and services sector using data compiled from an analysis of news articles on the UK Trade & Investment web-site. A total of 115 different foreign direct investment (FDI) projects were identified from January 2000 until the end of May 2004. Summary data on the date of investment, country of origin, type of investment and location of investment were compiled. A similar dataset - covering 101 FDI projects in the period April 1996 to September 1999 - was available for comparative purposes. This comparison is interesting because the period covered by the newer dataset was, broadly speaking, one of uncertainty (and possibly a downturn) in the global IT sector, whereas the period covered by the older dataset was one of strong growth for the sector both nationally and internationally.
6.2 Software FDI Patterns and Trends, 2000-2004 b. The number of software FDI projects attracted to the UK during the period was relatively evenly spread across the period, although there were significantly more projects in 2002 than in the other full years (Table 6.1). c. New developments were the dominant type of project during the period, accounting for 71% of all projects. There were also 20 acquisitions, 11 expansion projects and 2 joint ventures (Table 6.2).
d. London (40 projects), the South East (30) and East of England (10) attracted the largest number of software FDI projects (Table 6.3). Thus FDI in this period was acting to re inforce to existing dominance of the UK computer software and services industry by the Greater South East. Scotland and NI were the most successful regions in the rest of the country. The least successful regions were the South West and Wales – which
58
attracted only one project each during the period – and the North East – none at all. e. The United States was the major source of software FDI projects into the UK during 2000-04, accounting for 78% of all projects (Table 6.4). Canada (10 projects) was the other major source country. No other country accounted for more than 3 projects. Firms from mainland Europe and from the rest of the world each accounted for 11 software FDI projects in this period. f.
Within the US group, California was the leading State of origin (one third of all projects), followed by Massachusetts (16%). This is not unexpected since these States are home to the world renowned Silicon Valley-Bay Area and Boston-Route
128
technology
clusters.
Other
States
with
significant
technology clusters – such as Maryland (I-270 Tech Corridor), Colorado (Denver-Boulder), North Carolina (Research Triangle Park) – are the next best represented (Table 6.5).
g. The geographical distribution of projects by type of investment and country of origin is shown by F igures 6.1 and 6.2. In terms of type of investment, the distribution of new developments was quite similar to the overall pattern, with the majority being located in London and south-eastern England (Figure 6.1). Acquisitions - the other major investment type - were more common in the English North and Midlands where they accounted for half of all projects during 2000-04. h. In terms of country of origin, the distribution of US investments was quite similar to the overall pattern, with the majority (two-thirds) being located in London and South East England (Figure 6.2). Investments from the Rest of the World followed a broadly similar pattern. In contrast, Canadian and European investments were more likely to be located in the ‘North’ of the UK (about half of all projects in each case). i.
Before moving on, it should be stated that this analysis cannot comment on the number of jobs created by the various FDI projects, or their ‘quality’ and dynamic effects on the host region. This sort of commentary must be based on detailed firm-level inquiries. The potential for inter-regional variations in the size and quality of the projects analysed here should be noted.
59
Table 6.1: Number of Software FDI Projects Reported by Year, 2000-04 2004 (to end of May) Calendar year 2003 Calendar year 2002 Calendar year 2001 Calendar year 2000 Total
8 24 33 23 27 115
Table 6.4: Number of Software FDI Projects by County of Origin, 200004 USA Canada Ireland Germany Singapore Japan Australia France Malta Dubai South Africa India Italy Korea Israel Denmark Total
83 10 3
77.6% 8.7%
3 2 2 2 2 1 1 1 1 1 1 1 1 115
Aggregates: Europe Rest of World
11 11
9.6% 9.6%
Table 6.2: Number of Software FDI Projects by Type of Investment, 2000-04 New Development Acquisition Expansion Joint Venture Total
82 20 11 2 115
71.3% 17.4% 9.6% 1.7%
Table 6.5: Number of US Origin Software FDI Projects by State of Origin of Investing Firm, 2000-04 CA-California MA-Massachusetts MD-Maryland
27 13 5
CO-Colorado NC -North Carolina NY-New York WA-Washington FL-Florida IL-Illinois VA-Virginia TX-Texas WI-Wisconsin PA-Pennsylvania OH-Ohio GA-Georgia NV-Nevada CT-Connecticut MN-Minnesota NH-New Hampshire Total US origin
4 4 3 3 3 3 3 3 2 2 2 2 1 1 1 1 83
60
32.5% 15.7%
Table 6.3: Number of Software FDI Projects by Region of Location, 2000-04 London South East East of England Scotland Northern Ireland Yorkshire & Humber West Midlands North West East Midlands South West Wales North East Total
40 30 10 9 8 5 4 4 3 1 1 0 115
34.8% 26.1% 8.7% 7.8% 7.0% 4.3%
80 35
69.6% 30.4%
Aggregates Greater South East Rest of UK
Source for Tables 6.1-6.5: author’s analysis of UK Trade & Investment News Archive.
6.3 Comparison with Software FDI Patterns, 1996-99 j.
Comparison of the 2000-04 dataset with an equivalent dataset covering 1996-1999 reveals some interesting similarities and contrasts. First, the number of projects per year seems to have been slightly higher in 1998 and 1999 (before the global IT downturn) than in 2000 and 2001. However, the volume of projects has held up reasonably well in recent years, given the global uncertainty in the sector (compare Tables 6.1 and 6.6).
k. Second, there was a shift in the type of project received by the UK in 200004 compared to the earlier period. Notably there were fewer acquisitions during 2000-04 than in the (shorter) earlier period 1996-1999; but more new developments in the most recent periods (compare Tables 6.2 and 6.7). It is conceivable that multinational computer software and services firms became more conservative in their acquisition activity during the downturn period; 2002, perhaps the peak year of the downturn, saw the fewest acquisition events.
l.
Third, there was a dramatic shift in the regional location behaviour of foreign investors between the two periods. The Greater South East accounted for only 38% of all projects (n=38) attracted to the UK during 1996-99 but this increased to 70% (n=80) during 2000-04. In contrast, the number of projects attracted to the rest of the UK fell from 65 during 1996-99 to only 35 during 2000-04 (compare Tables 6.3 and 6.8). This point is well illustrated by the plight of Scotland and NI. During 1996-99 these two regions attracted 17 and 16 projects respectively, making them the two most successful regions at attracting software FDI projects. During 2000-04, however, they were only able to attract 9 and 8 projects respectively, compared to the 40 projects attracted by London and 30 by the South East. This reversal of fortunes may have something to do with a change in the character of the software FDI projects available to UK regions but further investigation would be required to confirm this.
m. Finally, the United States strengthened its dominant position as the leading country of origin for software FDI project; 78% of projects in 2000-04 compared to 63% in 1996-99. This shift reflects an increase in the actual number of US projects but also a reduction in the number of projects c oming to the UK from mainland Europe during the latter period (Tables 6.4 & 6.9).
61
Table 6.6: Number of Software FDI Projects Reported by Year, 1996-99 1999 (to end September) Calendar year 1998 Calendar year 1997 1996 (from April onwards) Total
34 32 27 8 101
Table 6.8: Number of Software FDI Projects by Region of Location, 1996-99 Scotland Northern Ireland South East East of England London West Midlands North West Yorkshire & Humber East Midlands South West North East Wales Multiple regions Total
17 16 14 13 11 7 7 4 3 4 2 0 3 101
16.8% 15.8% 13.9% 12.9% 10.9% 6.9% 6.9% 4.0% 3.0% 4.0% 2.0% 0.0% 3.0% 100%
38 63
37.6% 62.4%
Aggregates Greater South East Rest of UK
Table 6.7: Number of Software FDI Projects by Type of Investment, 1996-99 New Development Acquisition/merger Expansion Joint venture Relocation Total
53 29 13 4 2 101
52.5% 28.7% 12.9% 4.0% 2.0% 100%
Table 6.9: Number of Software FDI Projects by County of Origin, 199699 USA Canada Japan France Germany Netherlands Taiwan Austria Bermuda India Ireland Israel Sweden Total
63.5 8 7.5 7 4 3 2 1 1 1 1 1 1 101
62.9% 7.9%
17 12.5
16.8% 12.4%
Aggregates: Europe Rest of World
Note: one project was a JV (USA-Japan)
62
Source for Tables 6.6-6.9: author’s analysis of Invest in Britain Bureau News Archive and other news media and data sources. This data originally prepared for use in Crone (2000).
Figure 6.1: Location of Software FDI Projects, 2000-2004, by Type of Investment
Figure 6.2: Location of Software FDI Projects, 2000-2004, by Country of Origin
63
References
Coe N M (1996a) The Growth and Locational Dynamics of the UK Computer Services Industry, 1981-1996. Unpublished PhD Thesis, University of Durham. Coe N (1996b) Uneven development in the UK computer services industry since 1981, Area, Vol.28, 64-77. Coe N M (1997) US transnationals and the Irish software industry: assessing the nature, quality and stability of a new wave of foreign direct investment, European Urban and Regional Studies, Vol.4, No.3, 211- 230. Coe N M (1999) Emulating the Celtic tiger? A comparison of the software industries of Singapore and Ireland, Singapore Journal of Tropical Geography, Vol.20, No.1, 36-55. Crone M (2000) Computer Services and Software Employment in the UK and Ireland, 1991-97: An Exploration of Official Data with Particular Reference to Northern Ireland. NIERC Report Series No.16, Economic Research Institute of Northern Ireland, 22-24 Mount Charles, Belfast, BT7 1NZ. Crone M (2001) Software and Computer Services Employment in the UK Regions and Ireland, 1995-99: A Review of the New Annual Business Inquiry Data for Great Britain and Equivalent Data for Northern Ireland and Republic of Ireland. NIERC Report Series No.19, Economic Research Institute of Northern Ireland, 22-24 Mount Charles, Belfast, BT7 1NZ. Crone M (2002a) A Profile of the Irish Software Industry. Consultation draft published on-line at: www.qub.ac.uk/nierc . Crone (2002b) The Irish Indigenous Software Industry: Explaining the Development of a Knowledge-intensive Industry Cluster in a Less Favoured Region. Paper presented at the 42nd Congress of the European Regional Science Association, Dortmund, Germany, 27th–31s t August 2002. Johnston R (2001) One Million New Employees: Revisions to ONS Employment Data. Mimeo, Economic Research Institute of Northern Ireland, Belfast. McGuinness S, Bonner K (2000) A Study of the Northern Ireland Labour Market for IT Skills. Report Prepared by the Priority Skills Unit, Northern Ireland Economic Research Centre. Commissioned by the Northern Ireland Skills Taskforce. Published by Department of Higher and Further Education, Training and Employment.
64
McGuinness S, Doyle J (2003) Three Years on: Reassessing the Northern Ireland Labour Market for IT Skills: Final Report. Prepared by the Priority Skills Unit, Economic Research Institute of Northern Ireland. Commissioned by the Northern Ireland Skills Taskforce. Published by DEL (Department for Employment and Learning). ISBN 1871753465. O’Gorman C, O’Malley E, Mooney J (1997) The Irish Indigenous Software Industry An Application Of Porter’s Cluster Analysis. Research Series Paper No 3, National Economic and Social Council, Dublin. Ó Riain S (1999) Remaking the Developmental State: The Irish Software Industry in the Global Economy. Unpublished dissertation, Department of Sociology, University of California, Berkeley.
65
Appendix A: NASDAQ Stock Exchange Indices, 1997-2004
NASDAQ index historical data 450
COMP
SPX
400
NDX
350
300
250
200
150
100
50
07/30/2004
03/08/2004
10/13/2003
05/21/2003
12/26/2002
08/05/2002
03/13/2002
10/17/2001
05/21/2001
12/26/2000
08/03/2000
03/13/2000
10/19/1999
05/27/1999
01/04/1999
08/11/1998
03/19/1998
10/23/1997
0
Notes: COMP = NASDAQ Composite Index, a broad based Index measuring all NASDAQ domestic and international based common type stocks listed on The NASDAQ Stock Market; NDX = NASDAQ-100 Index, which includes 100 of the largest domestic and international non-financial companies listed on The Nasdaq Stock Market based on market capitalization. The Index reflects companies across major industry groups including computer hardware and software, telecommunications, retail/wholesale trade and biotechnology.; SPX = S&P 500 Index, widely regarded as the standard for measuring large-cap U.S. stock market performance, this popular index includes a representative sample of leading companies in leading industries. Source: http://dynamic.nasdaq.com/reference/IndexDescriptions.stm (accessed 02/08/04).
66
Appendix B: County Map of Great Britain
67
Appendix C i: Unitary & Local Authority District Map of the UK
Source: http://www.statistics.gov.uk/geography/downloads/UK_LADUACty.pdf
68
Zoom of midlands and northern England
Zoom of southeastern England
69
Appendix C ii: Local Authority District Map of Scotland
Source: Scottish Executive, Local Government in Scotland, Fact Sheet 12 (http://www.scotland.gov.uk/library3/localgov/fs12-02.asp)
70
Appendix C iii: Map of Northern Ireland NUTSIII regions
Notes: District Council Areas aggregate to five NUTSIII regions, colour-coded as follows:- Teal=Belfast; Magenta=Outer Belfast; Blue=East of NI; Red=North of NI; Green=West & South of NI . Source: DETI Statistics Research Branch, NI Census of Employment September 2003 “District Council Areas by Section” booklet, p. (published 22/12/04)
71
Appendix D: Location Map of British Provincial Cities
Note: the map shows the locations and boundaries of the Unitary Authorities covering Belfast and 18 leading British provincial cities outside the ‘Greater South East’. Regional boundaries and names are shown for reference purposes.
72
Appendix E: Membership of the ‘Western Arc’ in 2003 (& 2001) UALAD
Emp2003
LQ2003
LQRk03
Wokingham
9,093
758
1
1
Hart
4,691
742
2
1
Bracknell Forest
7,380
610
3
1
Runnymede
4,244
526
4
1
Windsor and Maidenhead
6,148
450
5
1
Surrey Heath
2,991
380
6
1
Stevenage
2,717
346
7
2
Woking
3,091
346
8
1
South Cambridgeshire
3,546
331
9
2
Spelthorne
2,400
321
10
1
West Berkshire
4,735
319
11
1
Rushmoor
2,836
319
12
1
Wycombe
4,673
297
13
1
Dacorum
3,500
294
14
2
Fylde
2,095
290
15
0
Warwick
3,705
285
16
0
South Buckinghamshire
1,518
277
17
1
St Albans
2,814
269
18
1
Mole Valley
1,904
268
19
1
Milton Keynes
6,479
266
20
0
Waverley
2,314
256
21
1
Slough
3,508
253
22
1
Elmbridge
2,473
253
23
1
Trafford
6,064
252
24
0
Hounslow
5,438
244
25
3
Basingstoke and Deane
3,313
234
26
3
Portsmouth
4,234
234
27
0
Kingston- upon-Thames
2,753
222
28
3
Solihull
4,342
221
29
0
South Oxfordshire
2,193
221
30
3
Reading
4,070
220
31
3
Telford and Wrekin
2,859
205
32
0
Richmond-upon-Thames
2,502
205
33
3
Hammersmith and Fulham
3,855
197
34
3
Havant
1,446
196
35
0
Horsham
1,736
195
36
3
Islington
5,574
193
37
0
Basildon
2,638
192
38
0
North Hertfordshire
1,740
187
39
3
Guildford
2,416
181
40
3
Cambridge
2,941
180
41
3
Hillingdon
5,642
180
42
3
Tandridge
1,082
180
43
0
Nottingham
5,947
178
44
0
Merton
2,170
177
45
3
937
177
46
3
Mendip
1,220
177
47
3
Chiltern
1 ,101
176
48
3
South Bedfordshire
1,272
172
49
Summary stats
Emp.
No.
% of UK
121,875
37
24.8%
1-3
66,813
17
13.6%
1
Three Rivers
Western Arc (contiguous >170) Inner core WA (contiguous >250)
Shading
Classification
3 Classification
Notes: Membership of Western Arc defined as all contiguous UALADs with LQ>170 (coded 1-3 in ‘classification’ column of table ). Inner Core of Western Arc defined as all contiguous UALADs with LQ>250 (coded 1 in ‘classification’ column). Noncontiguous UALADs (coded 0 in ‘classification’ column) with LQ>1 70 all shaded indigo on map (in table, names shaded yellow for North and orange for GSE ).
73
Location of the ‘Western Arc’ in 2003
Location of the ‘Western Arc’ in 2001 (for comparison)
74
