Local Labour Markets Within Wales

AN OVERVIEW OF THE WELSH LABOUR MARKET

Melanie K. Jones, Richard J. Jones and Peter J. Sloane

November 2002

Welsh Economy Labour Market Evaluation and Research Centre (WELMERC) Department of Economics University of Wales Swansea Swansea SA2 8PP

E-mail addresses: [email protected] [email protected] [email protected]

Abstract The aim of this paper is to paint a portrait of the Welsh labour market in the last decade. Several data sources are used to describe important aspects of the labour market namely employment, unemployment, inactivity, earnings, vacancies and education and training in Wales. As well as inter-temporal comparisons, the paper also compares the labour market in Wales to those in the rest of Britain and also considers variations across regions within the principality.

Acknowledgements The authors would like to thank David Blackaby for his useful comments and suggestions. Financial support from the European Social Fund is gratefully acknowledged.

Further copies of this paper can be found at the WELMERC website: www.swan.ac.uk/welmerc from December 2002.

1 AN OVERVIEW OF THE WELSH LABOUR MARKET

1. INTRODUCTION  Economic and Demographic Background  Education and Training

2. ACTIVITY AND INACTIVITY  Demographics  Intra-Regional Variations in Activity

3. EMPLOYMENT STRUCTURE  Industry Structure  Occupational Structure  Age Structure  Gender Divisions  Part-Time versus Full-Time  Self-Employment versus Employment  Intra-Regional Variations

4. EARNINGS  Unionisation  Intra-regional variation

5. VACANCIES

6. UNEMPLOYMENT  By Age  By Gender  Unemployment Duration  Comparisons with the rest of the UK  Intra-regional variation

7. SUMMARY AND CONCLUSIONS

2 AN OVERVIEW OF THE WELSH LABOUR MARKET

1. INTRODUCTION

The aim of this paper is to paint a portrait of the Welsh labour market over the last decade. Several data sources are used to describe important aspects of the labour market, namely employment, unemployment, inactivity, earnings, vacancies and education and training in Wales. As well as inter-temporal comparisons, the paper also compares the labour market in Wales to those in the rest of the UK and also considers variations across regions within the principality. In this paper we assume that there is an entity that we can describe as ‘the’ Welsh labour market. In practice, it may be more accurate to talk of Welsh labour markets given natural barriers between North and South Wales and close links between parts of East Wales and parts of England. The awarding of Objective One funding for West Wales and the Valleys has made such distinctions even more apparent since GDP1 per capita in this area was only 73% of the UK average in 2000 compared to 98% in East Wales. Similarly, there are important economic and labour market differences between urban and rural Wales. Midmore (1999) reports that in 1995 rural Wales (that is data drawn from predominantly rural Unitary Authorities) had a per capita GDP which was 92.7% of that for Wales and between 1985 and 1995 rural annual GDP growth was only 2.1% compared to 2.7% for Wales as a whole. In rural areas the demographic structure is also skewed towards the more elderly, partly contributing to a poor economic performance.

Economic and Demographic Background The fortunes of any labour market are inextricably linked to the developments in the product sector of the economy. Over the last decade the Welsh economy has been performing well by historical standards, but not as well as the rest of the UK. Per capita GDP in Wales is significantly lower than the UK average. In 1991, nominal GDP per head in Wales was 83.9% of the UK average of (£7,450). Despite a growth in total GDP from £21.53bn in 1991 to £30.69bn in 19992, faster growth in other regions meant that Wales fell further behind with per capita GDP, standing at 80.5% of the UK average in 1999 at £10,449. Of the British regions, only the North East of England had a lower GDP per head than Wales and in the UK Northern Ireland has a lower GDP per head than Wales. In 1998, the figure for Northern Ireland was 74.1 (Index UK =100) compared to 79.4 for Wales. Household disposable income in Wales was 88% of the UK average and individual consumption expenditure 87% of the UK average, higher than the figures for GDP. Part of the GDP gap is attributable to the relatively poor performance of the labour market in Wales since economic prosperity is determined, in part, by the number and ability of people who work in the economy.

1 Gross Domestic Product (GDP) is the most commonly used indicator of economic activity. It measures the total value of all goods and services produced over a given time period (usually a year) excluding net property income from abroad. 2 The 1999 figure is the latest figure available and is provisional.

3 Table one displays recently published figures from the 2001 Census, which show that out of a population of 2.9 million in Wales, nearly 60% are of working age, similar to the proportion in 1991, but less than the corresponding figure for England (61.5%).

Table 1: Population of Countries in the UK. Country Population Population of Working Rate Age 1991 2001 1991 2001 1991 2001 Wales 2,835,073 2,903,085 1,687,045 1,732,765 59.51% 59.69% Scotland 4,998,567 5,062,011 3,079,761 3,149,716 61.61% 62.22% England 47,055,204 49,138,831 28,822,690 30202736 61.25% 61.46% Source: Census (2001)

Cross border commuting to work also lowers the effective working population to some extent. For example, in 2001, 75,000 Welsh residents worked outside of Wales, whilst only 42,000 non-residents made the opposite journey. Calculations presented in the National Assembly’s ‘National Economic Development Strategy’ (2002) suggest that having a smaller proportion of the population of working age accounts for 2.5 out of the 20 percentage point gap between Wales and the UK average GDP.

Education and Training

Supply Side Problems The number of workers is an important determinant of a nation’s GDP. However, how much these individuals produce depends, in part, on their skills and abilities. Low levels of education and training can have negative implications for firms and individuals. Machin and Manning (1999) argue that globalisation and technical progress have significantly reduced the demand for unskilled labour. Thus, those people without formal qualifications are increasingly likely to have poorer labour market outcomes. Moreover, an absence of skilled workers can hamper the expansion of indigenous firms and deter other firms from relocating into Wales. In 1998, one in five firms who responded to a survey conducted by Future Skills Wales reported that they felt that ‘the skills of their employees are not high enough to meet their current business objectives’. Moreover, a quarter of these firms felt that this had restricted the development of their business. The same survey also found that 30% of firms had ‘hard to fill vacancies.’ 3 Employers suggested that half of these were difficult to fill because there were not enough suitably skilled people in the area (compared to 42% in the UK as a whole) and 20% because the applicants lacked the qualifications required. The absence of large companies is one reason why an area might have a deficient skills base, since these tend to be better training providers than small firms (Haughton and Peck, 1988). Supply side problems might be exacerbated by the migration of the more able workers from Wales to the South East of England (Forsythe, 1995).

Higher inactivity rates in Wales (discussed below) are partly attributable to greater participation rates in full-time education amongst 16 to 19 year olds. Throughout the last

3 Defined as vacancies that have been hard to fill in the 12 months prior to the survey.

4 decade Wales had higher percentage of 16-19 years in education than the rest of the UK. A study using data from England and Wales by Rice (1999), suggests that participation rates in further education, for both males and females, are positively related to the unemployment rate in the local labour market.

However, despite this higher participation rate, in 2002, a lower proportion of working aged people in Wales held qualifications at NVQ level 3 or above. Around 40% of working people in Wales held qualifications at this level compared to 48% in Scotland and 43% in England. This represents an improvement since 1993 when 29% of working aged people in Wales were qualified to NVQ level 3 or above. The disparity is smaller at NVQ level 4 where 27.1% of the working age population in Wales were qualified to this level compared to 27.6% in England and 32.6% in Scotland in 2002.4

Further analysis of the qualifications held by people of working age reveals that in 2002 the proportion with 'A' levels (or equivalent) and GCSE passes was relatively similar to England and Wales in 2002. In contrast, the percentage with a degree is lower in Wales (13.1% of males and 11.5% of females) than in England (17.3% and 14.6% respectively). Moreover, the proportion with no qualifications in Wales (17.8% of male and 20.8% of females) exceeds the corresponding figures in England (13.9% and 17.8%).

Table 2. Educational Attainment by Country.

Highest qualification for those of working age, LFS Spring Quarter 2002 Degree or Higher GCE A GCSE Other No Don't equivalent education Level or grades A-C qualifications qualification know equivalent or equivalent England Male 2,816,587 1,137,818 47,92,623 2,921,700 2,243,275 2,266,916 119,574 (17.3%) (7.0%) (29.4%) (17.9%) (13.8%) (13.9%) (0.7%) Female 2,143,309 1,273,694 2,592,235 4,028,784 2,145,635 2,450,233 77,593 (14.6%) (8.7%) (17.6%) (27.4%) (14.6%) (16.7%) (0.5%) Wales Male 121,911 70,171 275,165 169,045 115,212 165,589 12,174 (13.1%) (7.6%) (29.6%) (18.2%) (12.4%) (17.8%) (1.3%) Female 96,454 82,457 147,667 245,935 82,521 174,594 9,827 (11.5%) (9.8%) (17.6%) (29.3%) (9.8%) (20.8%) (1.2%) Scotland Male 267,927 159,717 573,559 227,489 141,269 199,315 7,097 (17.0%) (10.1%) (36.4%) (14.4%) (9.0%) (12.6%) (0.5%) Female 228,976 195,165 328,248 294,741 138,586 278,283 6,824 (15.6%) (13.3%) (22.3%) (20.0%) (9.4%) (18.9%) (0.5%) Northern Male 6,9798 32,847 179,727 94,216 33,784 121,111 Ireland (13.1%) (6.2%) (33.8%) (17.7%) (6.4%) (22.8%) Female 71,772 47,189 76,030 135,099 41,366 128,044 (14.4%) (9.4%) (15.2%) (27.0%) (8.3%) (25.6%) Source: QLFS, Data Archive.

Part of the explanation for the gap in formal qualifications between Wales and the rest of the UK may stem from a failure to acquire basic skills. The Basic Skills Agency reports a plethora of figures that highlight the short fall in the acquisition of basic skills in Wales. For example, the Agency reports that, at the end of primary school, 22% of Welsh

4 Source: Labour Force Survey.

5 children are below the required level in English and 26% in maths (Basic Skills Agency, 2001). Moreover, at key stage three, half way through secondary school, 38% of 14 year olds fail to attain the expected standard in English and the same proportion are below the target in maths.5 The cumulative effect of under attainment at school is highlighted by figures produced by the National Assembly, which estimates that over three-quarters of a million people in Wales have literacy and numeracy problems. Wales performs badly when compared to other industrialised countries. Figures from the International Literacy Survey reveal that a quarter of people of working age within Wales have poor basic skills compared to 22% in England, 12% in Germany and 7% in Sweden.6 A survey by the Basic Skills Agency conducted between 1996 and 1999 revealed variations in the basic skills problem within Wales. Areas with high unemployment and high levels of social housing such as Blaenau Gwent, Torfaen and Merthyr Tydfil, also have the most acute basic skills problem. This poses a question over the direction of causation. Do people with low skills cluster together in low cost housing in high unemployment areas or do people have low skills because they are from these areas and perhaps lack the opportunities to acquire these skills?

Some information on the distribution of broad skills is available in the Work Skills in Britain Study (Felstead et al. 2002). This estimates a required qualifications index based on the qualification required to get the current job held by respondents to the survey. In Wales this was 1.95 compared to 2.41 in London and 2.30 in the South East. Only the East Midlands and Eastern Regions had (slightly) lower figures than Wales. A training time index based on the time taken in training for the type of work currently done showed a lower figure in Wales than any other region apart from the East Midlands, though the learning type index based on time taken to become fully proficient at the job suggests that the complexity of job is not very different in Wales than elsewhere. Data from the same survey on the regional distribution of generic skills shows that Wales is above average in the requirement for physical and horizontal communication skills but below average in high level communication skills.

Table 3. NVQ Level 3 Qualifications by Objective One Area. Number of working age qualified to NVQ level 3 or above in Wales

% Change % Change Area 1998 1999 2000 2001 2002 1998-2000 2000-2002 Number Objective (‘000s) 371 392 382 382 413 2.96 8.12 One Proportion 33.8 35.2 34.6 34.9 37.0 2.37 6.94 Number Objective (‘000s) 267 245 260 289 280 -2.62 7.69 Three Proportion 41.8 39 40.3 43.4 43.1 -3.59 6.95 Source: QLFS, Nomis. All data refer to the quarter ending in February of each year.

5 National Curriculum Assesment Results in Wales – Key Stages 2 and 3. The National Assembly for Wales 2001, reported in 6 Adult Literacy in Britain: A Survey of Adults aged 16-65 in Britain part of the International Adult Literacy Survey (IALS). ONS 1997.

6 Table 4. NVQ Level 4 Qualifications by Objective One Area. Number of working age qualified to NVQ level 4 or above in Wales

% Change % Change Area 1998 1999 2000 2001 2002 1998-2000 2000-2002 Number Objective (‘000s) 191 204 197 209 207 3.14 5.08 One Proportion 17.4 18.3 17.8 19.1 18.6 2.30 4.49 Number Objective (‘000s) 145 140 155 182 153 6.90 -1.29 Three Proportion 22.7 22.4 24.0 27.3 23.6 5.73 -1.67 Source: QLFS, Nomis. All data refer to the quarter ending in February of each year.

Tables 3 and 4 show the qualification attainments in areas with different European Objective funding status. Between 1998 and 2002, educational attainments, for those of working age are lower at NVQ level 3 and 4 in the Welsh Objective One area. The proportions with NVQ level 3 and 4 have increased over the period, and the rate of increase in the Objective One area (since 2000) has been greater, closing the gap on the Objective Three area.

Training Provision In August 2002, 11.8% of working age people in Wales were receiving job-related training,7 lower than the proportion in England (12.5%) but higher than the rate in Scotland (10.3%). There are differences in the provision for males and females. In Wales, 14.4% of working age women receive job related training. This is higher than the rates in England (13.2%) and Scotland (11.3%). In contrast, less than 10% of working age males in Wales receive job related training (9.5%), the same rate as in Scotland but lower than the rate in England (11.9%).

2. ACTIVITY AND INACTIVITY

Another important factor determining the output of an economy is the number of working age people who actually participate in the labour market. Using the ILO definitions, people are classified as being economically active if they are either in employment or unemployed. In contrast, those who are retired, those in full-time education, those who are unable to work due to sickness or disability and those who choose not to work, for example because they are looking after family or have become discouraged from searching for work, are classified as economically inactive. This suggests that the economically inactive can be broadly divided into two groups. First those for whom inactivity is not a chosen state but are forced to become inactive. For example, there are 440,00 disabled people in Wales, accounting for nearly 23% of the working population, compared to an average of only 18.2% for Great Britain as a whole.8

The second group consists of those for whom inactivity is the outcome of rational economic decision-making. Job search theory suggests that high unemployment and the

7 Defined as having received some form of job related training in the previous month. 8 Inactivity rates are expressed as a percentage of the total population of interest, and not the working population, as is unemployment.

7 associated low employment probabilities for unemployed individuals reduces job search incentives and thus search intensity (see Devine and Keifer (1991) for the seminal survey of job search theory). In the extreme case, the unemployed individual may rationally choose to withdraw from the labour market and become inactive, when the expected value of search activity is lower than the costs associated with undertaking that search. This may be particularly true for older workers who have less time to reap the benefits of job search than young workers.9 Evidence presented by Manning and Burdett (1996) suggests that this does happen in the UK.

One of the reasons why Wales has a lower GDP per head than the rest of the UK, is that a higher proportion of working age people in Wales are inactive. Table 5 displays information that shows the extent of inactivity in Wales compared to England and Scotland for both males and females. Using the strong assumption that those who are inactive are potentially as productive as those who are employed, the National Assembly estimates that higher inactivity rates account for 5.3 percentage points of the GDP gap.10 Less than one percentage point of the GDP gap is ascribed to differences in the unemployment rate. Finally, 11.9% of the GDP gap is attributed to differences in productivity between Wales and the rest of the UK. One component of this productivity gap might be due to the structure of employment in Wales.

Table 5: Activity Rates Country Total Activity Rates Male Activity Rates Female Activity Rates (%) (%) (%)

August August August August August August 1992 2002 1992 2002 1992 2002 Wales 73.6 74.7 80.6 80.3 65.8 68.5 Scotland 79.2 79.6 86.6 83.3 71.3 75.7 England 80.1 79.8 87.9 85.2 71.6 73.8 Source: QLFS, Nomis

Over the decade, total inactivity rates remained relatively stable, a pattern replicated in the rest of Britain. Inactivity rates differ by standard statistical region; in 1991 inactivity was highest in Wales, at 26% and lowest in the South East at 17.5%. In 2002, the rates were 25.3% and 16.6% respectively, showing no tendency for convergence, unlike regional unemployment. The regions of high inactivity and unemployment tend to coincide, suggesting inactivity is not only determined by supply side factors (Nickell, 2001).

Using the annual Local Labour Force Survey, inactivity can be decomposed into its various components. In 2001, Wales had the highest proportion of inactive workers, who ‘do not want to work’, at nearly 75% compared with 73% for England and 65% for Scotland. O’Leary et al. (2002) note that one of the reasons for high levels of inactivity

9 More specifically, for older workers the discounted value of a wage received until retirement will be lower than the same wage received until retirement for younger workers. Therefore, other things being equal, the older worker is more likely to find that the cost of search exceeds the benefits of search. 10 It maybe that those who are inactive are potentially less productive, the resulting lack of opportunities available to them is part of the reason why they withdraw from the labour market.

8 in Wales is the relatively high percentage who describe themselves as long-term sick. Between the late-1970s and the mid-1990’s the number of incapacity/invalidity benefit claimants more than doubled in Wales. In 1991, this figure was a third higher than in England. The higher Welsh figures may, in part, result from limiting long-term illness associated with concentrations of heavy industry. More surprisingly, the proportion inactive because they are looking after the family home is lower for Wales at 6.1% compared to 8.1% in England and 9.1% in Scotland.

Demographics and Inactivity There are important gender differences in the dynamics of inactivity over time. Between 1991 and 2002 male inactivity rose in all UK regions, though by the smallest percentage in Wales. Female inactivity fell during the decade, consistent with increased participation, narrowing the gap in activity rates between genders. Blackaby et al. (2001) suggest increased female earnings, improvements in household technology and changes in work preferences contribute to increasing female participation. Both male and female inactivity rates remain above the UK average in August 2002, at 19.7% and 31.5% respectively.

To analyse the effect of age on inactivity three age groups, 16 to 24, 25 to 49 and 50 years old to retirement age are used. Inactivity in Wales exceeds the British average for all these groups. For the age group 16-24, inactivity has risen by 12% in Wales, whereas for Britain the rate remains fairly stable. Thus the gap in youth inactivity between Wales and England has increased over time. In contrast, inactivity between the ages 25-49 has fallen within Wales to 17.1% (August 2002), bringing it closer to the rate in England of 15%, which has remained relatively stable over the period. Inactivity rates exhibit their traditional U shape, being higher for the 16 to 24 year olds and 50-year-old to retirement ages. In Wales rates for the 50 to retirement group exceed the 16-24 group throughout the period, whereas in England the two rates are more similar. Although declining over the period inactivity in Wales for this group is 37.7% in August 2002, substantially greater than 29% for the UK as a whole. High inactivity in the 50 year old to retirement group is consistent with job search theory.

Intra-regional Variations in Inactivity The level of inactivity varies between the unitary authorities of Wales. In May 1994, the percentage of working age classified as inactive was highest in Rhondda, Cynon, Taff at 37.1% and lowest in Powys at 14.6%. In May 2002 inactivity varied between 13.6% in the Vale of Glamorgan and 41.2% in Ceredigion. Thus, intra-regional inactivity has shown no tendency to converge over the decade. Although for Wales as a whole total inactivity remained relatively stable through the ninties, unitary authorities experience greater variability over time. For example, Torfaen exhibited a 40% rise and the Vale of Glamorgan a 40% fall between May 1994 and May 2002.

When areas are grouped by Objective One and Objective Three status, the level of activity is consistently higher in the Objective Three area. Between 1996 and 2000, the level of economic activity rose in the Objective Three area, but fell slightly in the

9 Objective One area. In contrast, between 2000 and 2002 the number economically active fell in the Objective Three area, but not in the Objective One area.

Table 6. Economic Activity in Different European Objective Funding Areas. Economic Activity in Wales % % Change Change 1996- 2000- Area 1996 1997 1998 1999 2000 2001 2002 2000 2002 Number Objective (‘000s) 797 800 790 816 791 777 798 -0.75 0.88 One Rate 72.4 72.7 72.0 73.2 71.6 71.0 71.5 -1.10 -0.14 Number Objective (’000s) 473 491 479 481 510 520 489 7.82 -4.12 Three Rate 75.9 78 75.0 76.7 79.0 78.0 75.1 4.08 -4.94 Source: QLFS, Nomis. All data refer to the quarter ending in February of each year.

3. EMPLOYMENT STRUCTURE

Industry Structure Since the late 1970’s, one of the defining characteristics of the Welsh economy has been a structural change away from heavy industries such as coalmining and steel production toward a more service-based economy. As in the rest of Britain, the de-industrialization process has continued through the 1990’s with an expansion of employment in all the broad economic sectors apart from manufacturing.11 There have been several notable examples of large-scale job losses across Wales such as the closure of BP Llandarcy, 12 the downsizing of steel-making plants, and the closure of several suppliers of Marks and Spencer e.g. Dewhirst in Fishguard and FII in Bridgend. Over the last 30 years there have been over 200,000 job losses in declining industries in Wales as a whole.

As figure one indicates, in 1992, almost a quarter of million people, one-in-five of the working population, were employed in manufacturing. By 2002, this had fallen to 210,000 people or 16% of the workforce. Conversely, over the same period, the number employed in the service sector increased by 130,000 to 913,000. The largest part of the expansion of the service sector has been in public administration where 88,000 more people are employed now than a decade ago. The fall in manufacturing and rise in services, between 1992-2002, indicates a continuation of earlier trends identified by Cameron et al. (2002). The rising property market has helped to fuel an increase in the number of people employed in the construction sector from 86,000 in 1992 to 104,000 in 2002.

In comparison with the rest of Great Britain, Wales has a higher proportion of the workforce employed in public administration (30.6% compared to 24.1% in England and 28.1% in Scotland). In contrast, despite a rise of 26,000 employees over the decade, the proportion in banking and finance, a high wage industry, remains low at 9.3%, compared

11 SOC single digit codes. 12 See An Economic and Social Assessment of the Closure of BP Llandarcy: Summary Report, University of Wales Swansea, Swansea, 1999 for a discussion.

10 to the British average of 16.1%. The problems afflicting agriculture in UK over the last decade have been well documented. This is reflected in a decline in the number of people employed in agriculture in Wales, which fell from 64,589 in 1991 to 55,737 in 200013. However, the number of farmers, partners and directors only fell by just over a thousand.

Figure 1

Employment Structure In Wales 1993 - 2002

100% Other Industry Sectors

90% Other Services

80% Public Administration 70% e r

a 60% h Banking & Finance S

e 50% Transport & m

y Communication o l p

E Distribution

30% Construction 20%

10% Manufacturing

0% 1993 1994 1995 1996 1997 1998 1999 2000 2001 Year

Source: QLFS, Nomis.

A notable feature of the Welsh economy has been its ability to attract investment from overseas. As Roberts (1994) notes, foreign firms have been investing for many years but it is only recently that these firms have become a significant source of employment and have been some of the most successful firms in Wales.14 At the turn of millennium, nearly 74,000 people were employed in the 356 foreign owned manufacturing plants in Wales15, representing a third of the manufacturing workforce. Similar numbers were employed in tourism-related industries (77,300).16

Figures published by the National Assembly17 show that in 2000 just under 11% of the workforce in Wales (135,000 people), were employed in hi-tech industries, with 8.0% in

13 National Assembly for Wales, Agriculture, Fishing and Forestry: Labour and Earnings. 14 Munday and Peel (1997) for example, find the performance of foreign firms in Wales better than their domestic counterparts, in terms of sales, assets and employment, during the 1990s recession. 15 Source: Welsh Local Area Statistics: Table 7.7 Employment in overseas owned manufacturing plants, 2000 (a). 16 Source: Welsh Local Area Statistics: Table 7.5 Employee jobs in tourism-related industries, September 1999 (a). 17 Source: Employment within the Hi-Tech industries within the EU, 2000, National Assembly for Wales Statistical Bulletin 2000.

11 hi-tech manufacturing and 2.9% in hi-tech services. This was compared with a total of 14.1% in Germany, 12.6 % in Sweden and the UK average of 11.6%. The lowest concentrations in the EU were in Greece (3.4 %) and Portugal (4.5%).

Table 7: The Industrial Structure of Employment in Great Britain August 1992 August 1996 August 2000 August 2002 d d d d d d d d s s s s n n n n n n n n e e e e a a a a a a a a l l l l l l l l l l l l t t t t a a a a g g g g o o o o n n n n c c c c W W W W E E E E S S S S

% all of employed workers who work in: manufacturing 21.2 20.5 19.3 19.2 21.8 16.9 16.9 17.3 14.6 15.8 16.3 13.7 construction 6.7 7.1 8.0 6.9 6.6 7.9 7.0 8.1 7.9 7.2 8.1 7.0 distribution 20.3 20.5 18.7 20.0 22.6 20.3 19.7 19.1 19.4 19.7 19.9 19.5 transport and communication 6.4 5.4 6.1 6.4 4.8 6.4 7.2 6.6 7.0 7.2 5.0 6.5 banking and finance 11.9 8.2 8.8 14.7 7.7 11.1 16.5 10.5 13.1 16.6 9.3 14.5 public administration - - - 23.8 26.7 25.9 24.2 28.3 26.4 25 30.6 28.3 other services 29.1 32.4 31.7 6.0 5.1 6.1 6.2 5.8 5.9 6.1 6.1 5.6

Occupational Structure The occupational structure of employment in Wales has also changed in the last decade as evidenced by the information in tables 8a and 8b. There are four thousand fewer workers in unskilled occupations and eighteen thousand less in clerical occupations; though the percentage share of clerical jobs has risen slightly. In contrast, there are 38,000 more professionals and 28,000 more associate professionals than a decade ago. A corollary of this is that by 2002, Wales had a higher proportion of both professional workers and unskilled than the rest of the UK, but a smaller proportion of employees in managerial or senior administrative roles.

12 Table8a: Occupational Structure of Employment in the UK (%s). August 1992 August 1996 August 2000 Occupational England Scotland Wales England Scotland Wales England Scotland Wales Group (SOC90) Managers & 15.6 12.0 15.7 16.1 12.8 15.4 16.5 14.2 13.5 Administrators Professional 9.9 9.8 8.8 10.5 10.3 8.6 10.9 10.3 10.6 Associate Professional & 9.1 8.7 7.8 9.7 9.9 8.7 10.9 10.1 9.3 Technical Clerical 15.6 14.1 14.0 15.0 14.4 12.6 14.7 14.4 14.4 Craft & Related 13.8 14.9 13.3 12.3 13.0 13.2 11.6 12.1 12.4 Personal & Protective 9.7 10.2 10.9 10.7 11.7 11.6 10.8 12.1 11.1 Services Selling 7.8 8.1 7.4 7.8 8.3 8.6 8.2 8.7 8.2 Plant & Machine 9.2 10.1 11.1 9.4 9.5 11.1 8.6 8.8 11.0 Operative Other 8.7 11.3 10.4 8.0 9.6 9.6 7.4 9.0 9.1 occupations Unskilled 5.4 6.8 6.2 4.9 5.5 6.3 4.5 5.0 5.3 Source: QLFS, Nomis.

Table 8b: Occupational Structure of Employment in the UK (%s). August 2002 Occupational Group (SOC2000) England Scotland Wales Managers & Senior Officials 14.7 12.4 11.2 Professional 11.5 12.1 12.2 Associate professional & technical 13.7 12.8 12.6 Administrative and secretarial 13.2 13.2 11.2 Skilled trades 11.5 12.2 13.5 Personal service 7.1 7.2 8.0 Sales and customer service 7.7 8.2 9.1 Process, plant & machine operatives 8.2 8.6 9.6 Employment in elementary 12.1 12.9 12.5 Source: QLFS, Nomis.

The Age Structure of the Workforce In 1992, 60.6% of people aged between 16-24 in Wales were employed; this was lower than the corresponding figures for England (61.6%) and Scotland (63.9%). This difference is partly due to the higher participation rates in further education in Wales noted above. Over the decade, the employment rate in this group fell more quickly in Wales than the rest of the UK, resulting in wider differences by 2001, when the corresponding figures were 52.7% for Wales, 61.8% for England and 62.6% for Scotland.18

During the 1990’s, the proportion employed in the 25-49 years old group increased in all countries, but the regional differences remained in 2001. In 1992, the proportion of those

18 All figures refer to the quarter ending in May of that year.

13 aged 25-49 years old who were employed, was 74.5% in Wales, 77.6% in England and 77.3% in Scotland. In 2001 the corresponding figures were 78.3%, 81.6% and 80.3%.

The proportion of people aged over 50 years in employment is far less than the other age groups. However, over the period, this proportion rose, by around 15% in all areas. Again, Wales had the lowest figures throughout the decade at 26.8% in 1992 and 30.6% in 2001, compared with 31.6% and 36.7% for England and 31% and 32.8% in Scotland. Even though employment rates were lower in this group, the concentration of population over fifty in Wales meant that, in 2001, Wales had the highest proportion of its workforce aged over 50 years. Moreover, population projections suggest that by 2005 approximately 50% of the working age population in Wales will be over the age of 40.

Gender Differences in Employment The proportion of working age females that were employed was lower than males, for all countries over the entire decade. In Wales in 1992, 71.6% of males were employed and 62.5% of females, the corresponding figures for 2001 were 73.6% and 62.8%. All these figures lie below the proportions in England, 76.8% and 65.8% in 1992 and 79.7% and 70% in 2001 for males and females respectively. The proportion of females in total employment has been relatively stable over the period at around 44% for Wales and for England and higher in Scotland. In the first quarter of 2002, 18.8% of working males were employed in the public sector and for working females the figure was 39.4%.

Similar gender differences exist in all UK regions but employment in the public sector 19 is higher in Wales than England, but lower than either Scotland or Northern Ireland. In the same quarter in 1994 similar figures for Wales are 21.1% for males and 38.5% for females, indicating an increase in private sector employment for males and the converse for females, since that time.

Part-time/Full-time The ratio of full-time to part-time workers, for both the UK and Wales has been fairly consistent throughout the decade at three to one. However substantial gender differences remained in 2001, 51% of females worked part-time compared to only 15% of males. 20 O’Leary et al. (2002) suggest that increased demand in female dominated sectors has resulted in the growth in the number of part-time jobs, to attract women with family commitments.

The (mean) average number of hours of work has also remained relatively constant over the period. In 2002, the average number of hours worked in Wales including overtime is 39.5, just slightly less than the figure for England, 39.7 hours per week.21 Self-Employment/Employment It has been suggested that one of the reasons for the lower GDP and employment rate in Wales is the lack of entrepreneurial spirit. Wales has a lower business start up rate than the rest of the UK and the European average. Recently there have been a number of 19 As opposed to public administration, which was discussed in the employment section. The public sector being the wider definition since it also includes sectors such as health and education. 20 Labour Market Trends August 2001. 21 Source NES.

14 initiatives to encourage more people in Wales to start their own business. For example, the Welsh Assembly has launched the Entrepreneurship Action Plan that aims to increase the number of individuals starting a new business and improve the growth of existing businesses. However, as a proportion of total employment, the rate of self-employment in Wales is similar to that in England at around 11.7%, and higher than the rate in Scotland (9.7%). A decade ago Wales was the leader in the UK with 13.9% of all employed workers being self-employed compared to 12.7% in England and 9.4% in Scotland. There is a gender inequality in self-employment, with working men twice as likely to work for themselves as working women throughout the UK.

Intra-Regional Variations in Employment The proportion of working age in employment varies by unitary authority due to both differences in activity and unemployment rates. In 1993, employment ranges from only 59.5% in Bridgend to 80.6% in Powys; by 2001 Ceredigion had the lowest employment rate at 55.6% and Monmouthshire the highest at 77.3%. Employment in the Objective Three area, in both 1993 (70.6%) and 2001 (73%), is higher than similar figures for the Objective One area (64.5% and 67.3%).22

In November 1993, Neath and Port Talbot, had the lowest ratio of self-employed workers to total workers in Wales at 3%. At the other extreme, 43.9% of workers in Powys were self-employed. By May 2002, Bridgend had the lowest proportion of self-employed workers (4.4%) and Powys had still the highest with 27.4%. Between 1993 and 2000 self-employment accounted for a larger proportion of employment in the Objective One area than Objective Three area. By August 2002, the proportion in self-employment was 13% in the Objective Three area and 11% in the Objective One area.

The amount of full-time employment also varied from 83.2% in Caerphilly to 67% in Conwy in 1993. By 2001, the growth in full-time employment in Conwy meant the rate rose to 82.1%, the highest of any unitary authority in Wales; in contrast a decline in full- time Pembrokeshire resulted in only 67% being employed full-time. Employment is comprised of a slightly greater amount of full-time work in the Objective Three area, and this remains relatively stable over time. In 2001, the full-time proportions were 75% in the Objective Three area and 73.4% in the Objective One area.

Intra-Regional Variations in Industry Structure When considering Wales by European Objective status, differences emerge in both the industrial and occupational structure of employment. The Objective One area had a lower proportion in banking and finance, but the growth in financial services in the Objective One area exceeded the Objective Three area, reducing the difference by 2001. A similar convergence in the proportion employed in manufacturing exists, with a greater decline in the Objective One area. In contrast, employment in public administration increased more quickly in the Objective One area. By 2001, 30.3% were employed in this sector, compared to 27.3% in the Objective Three area.

Table 9: The Industrial Structure of Employment in Different European Funding Regions in Wales

22 All data refer to the quarter ending in November.

15 % of all employed by industry

Objective One Objective Three

d d n n a a

t t g g r r n n o o

i i p p g g s s k k n n n n s s n n n n e e i i n n i i o o n n c c o o r r i i a a i i i i a a m m t t t t u u r r v v B B d c d c t t r u r u T T e e a a c c u u e e

b b r r c c n a n a i i s s t t c c f f

r r n n o o i i s s t t r r i i l l u u a a t t s n s n e e b b n n i i n n a a o o i h i h a a u u c c t t D D i i F F C C P P

M M O O n n d d u u n n a a m m m m o o c c 1993 Number (‘000s) 164 62 150 33 49 185 31 87 33 82 27 47 121 26 % 22.5 8.5 20.5 4.5 6.7 25.4 4.2 19.3 7.2 18.1 6.0 10.4 26.7 5.7 2001 Number (‘000s) 132 61 156 43 70 232 42 88 37 101 29 55 136 36 % 17.2 7.9 20.3 5.6 9.1 30.3 5.4 17.6 7.3 20.1 5.9 11.1 27.3 7.2

There are important differences in industrial structures between rural and urban areas. Rural areas are heavily dependent on a high level of public sector employment, though at predominantly lower grade in terms of pay and occupation. Midmore et al. (1994) argue that a number of measures are required to reverse the decline in rural employment including diversification through linked food processing activities, increases in areas of forestation, more rural tourism and more manufacturing.

Intra-Regional Variations Occupational Structure

Table 9 Occupation Structure in Objective One and Objective Three Areas Objective One Objective Three 1993 2000 1993 2000 Number % of total Number % of total Number % of total Number % of total (‘000s) employment (‘000s) employment (‘000s) employment (‘000s) employment Managerial and 101 Administration 13.8 97 12.7 68 15.1 74 14.8 Professional 54 7.4 83 10.9 58 12.8 70 13.9 Associate Professional and Technical 55 7.6 65 8.6 36 7.9 55 11.0 Clerical 94 12.9 105 13.9 64 14.2 70 13.9 Craft and related 106 14.5 102 13.5 56 12.4 54 10.7 Personal and Protective 76 10.4 86 11.3 45 10.0 53 10.5 Selling 59 8.1 59 7.8 36 8.0 39 7.8 Plant and machine operatives 94 12.9 87 11.5 45 9.9 50 10 Other 85 11.6 74 9.8 42 9.3 36 7.1 Unskilled 48 6.6 45 6.0 21 4.6 19 3.8

16 The Objective Three area had greater proportions employed in managerial, professional and associate professional and technical groups. In contrast, employment is concentrated more in craft, plant and machine operative and unskilled groups in the Objective One area.

4. EARNINGS

This discussion of earnings is based on the New Earnings Survey (NES) conducted by the ONS. The survey is based on a 1% sample of employees who are members of a Pay-As- You-Earn income tax scheme. The survey covers Great Britain and data are available for Wales as a whole. Since 1996, data are available at the sub-regional level, for local authority districts in Wales.

In 2001, employees in Wales had the lowest average (mean) gross weekly earnings for full-time workers in Great Britain, at £382 compared with £452 for England and £405 for Scotland. The low average is the result both of a high proportion at the bottom end of the pay distribution and a low proportion at the upper end of the pay distribution. In Wales, 26.5% of workers earn less than £250 per week and only 27% earn more than £460, comparable figures for England being 19.8% and 34.8%, thus, indicating fewer in the lowest pay group and more in the highest. Between 1997 and 2001, average earnings rose in Great Britain (by 20.9%) but the growth was slowest in Wales (15.7%). Therefore, average earnings in Wales exhibited no tendency to converge with the UK. This is consistent with the findings of Cameron et al. (2002), who for the period 1975 and 1995, identify a decline in relative earnings in Wales. As suggested above, low earnings in Wales may explain some of the lower activity rates since low wages mean that there is less incentive to participate in the labour market.

Although nominal wages are lowest in Wales, regional price differences may explain part of this differential. In January 2002, Economic Trends published average price level differences (in 2000) between London and UK regions. Only the North East had lower average prices than Wales, which were 3.8% below the UK average23. The nominal average wages discussed above can be adjusted using the relative price levels to calculate real wages. In 2001, average nominal earnings in Wales were £382, 14% below the UK average of £442. When price level differences are taken into account the real earnings differential is reduced to 10%.

The 90/10 earnings differential is used to measure the skewness of the earnings distribution and thus earnings inequality. The 90/10 earnings differential expresses the earnings figure for the top 10% of workers divided by the figure for the lowest 10% of earnings. In 1997, the figure was 3.19 for Wales and 3.37 for Great Britain, similar figures for 2002 were 3.14 and 3.49. Thus, Wales exhibits lower earnings inequality than Great Britain. Over time, earnings inequality in Wales has narrowed slightly, in contrast to the widening earnings inequality in Great Britain for the period 1975-1996 identified by Dickey (2001). Figure 2.

23 The basket of goods considered included all products including housing rents.

17 Average Weekly Earnings In Great Britain 1997-2001 - All Workers

e 400 e w

r e p 380 £

320 1997 1998 1999 2000 2001 Year

Great Britain England Wales Scotland

Source: NES, Office of National Statistics.

The differing impact of the minimum wage across regions may have contributed to the above trends. Low paid workers are over represented in Wales and thus the effect on the distribution of earnings may be greater in these areas.24 Robinson (2001) examines the regional impact of the national minimum wage and finds the proportion of low paid men fell by 64% in Wales between 1998 and 1999, similarly by 45% for females, exceeding national effects, both more than in the rest of Britain.

24 In 1997, the earnings of the lowest 10% of Welsh workers was less than similar figures in England and Scotland. Only Northern Ireland and North East England are found to have a higher incidence of low pay than Wales by the Low Pay Commission in 1997. It is clearly possible that reductions in employment may coexist with increasing wages in Wales.

18 A number of factors may contribute to Wales’ lower average earnings than the rest of the UK. One set of explanations focus on the occupational structure of the Welsh labour market, notably the relative dearth of well-paying, senior, managerial and administrative posts compared to the rest of the UK noted above. One of the major sources of new jobs in Wales has been the establishment of call centres. These have been cited as a source of low pay, with average wages in these centres being £11,100 per annum (TUC, 2001). Moreover, a reliance on public sector jobs that are associated with lower wages, may also depress wages. Public sector employment is associated with a compressed earnings distribution, Blackaby et al. (1999) find public sector workers experience increased premiums in the lower tail of the wage distribution and underpayment at the opposite end of the earnings distribution. The concentration of Welsh workers, especially females, in public sector employment may contribute to regional and gender earnings differentials.

Cameron et al. (2002) argue that the decline in full-time men’s earnings in Wales relative to the rest of Britain over the period 1975 to 1994 was caused by long-run factors that are unlikely to reverse themselves. That is it was an equilibrium phenomenon caused by a tendency for jobs to disappear in well-paid industries and be replaced by jobs in lower-paid industries.

Human capital theory suggests that the wages of an individual depend on their educational attainments and accumulated work experience. It was noted above that a lower proportion of individuals in Wales attain NVQ level three or above, than in the rest of the UK. Moreover, individuals who do have high qualifications are often attracted out of Wales to places where there are more suitable employment opportunities. Wales, also, has lower rates of functional literacy and numeracy than parts of England. The theory of compensating wage differentials predicts that low wages will coexist with attractive ‘compensating’ job features, such as a pleasant working environment. Despite earning relatively less than workers in the rest of the UK, there is some evidence to suggest that workers in Wales are happier in their jobs than in other parts of the UK

Average weekly earnings for full-time workers differ considerably by gender throughout the UK, with the female average consistently below the male (see figure 5). In Wales in 2002, average female earnings were £345 per week, £88 less than the male average. The 25% raw wage gap is lower than either Scotland or England. Further examination of the Welsh gender pay gap by Blackaby et al. (2001) suggests 25%-50% of the raw differential could be due to discrimination. Some of the pay gap is explained by an under-representation of women at senior levels (see Blackaby et al.1999).

The average weekly earnings of full-time non-manuals exceed manuals throughout the UK. In Wales, the differential is smaller than elsewhere, with average manual earnings at £326 for manuals and £440 for non-manuals.

Intra-Regional Variation In 1997, average earnings were lowest in Conwy at £284 and highest in Neath and Port Talbot at £351. In 2001, Flintshire had the highest mean average earnings at £421, just

19 above Neath and Port Talbot (£417) and Conwy still exhibited the lowest average at £326. The high average wage in Neath and Port Talbot in part reflects the concentration of employment in manufacturing. The decline of manufacturing with relatively high paid jobs (e.g. Corus) is a concern to the whole economy, but since the effects are spatially concentrated, sub-regional disparities may widen. Analysis of the Objective One area also shows poor performance in terms of average earnings. In 1999, the average earnings in the non-Objective One area were £313, substantially higher than the £289 average in the Objective One area. Although average earnings have risen, in 2002 the gap remains with similar figures of £356 and £320 respectively.25 The lower proportion qualified to NVQ levels 3 and 4 (tables 3 and 4) in the Objective One area may explain part of this wage gap.

Figure 3.

Average Weekly Earnings In Wales by Gender 1990-2000

400 ) £ (

g 350 n i n r

a 300 E

r 250 e v A 200

150 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year

Persons Males Females Source: Welsh Digest of Local Area Statistics.

Unionisation It has long been suggested that trade unions may influence earnings in (Hildreth 1999). Workers in Wales are more likely to be members of a trade union than workers in any other region in the country. In 2001, the percentage of males and females employed that belonged to a trade union or staff association were 35.6% and 35.3% respectively. 26 Comparable figures are 25.2% and 25.6% for England and 31.0% and 34.0% for Scotland. Union membership has declined over time, particularly for males. In 1994, corresponding figures for Wales were 42.4% and 38.7%. The effect on earnings may depend on union recognition rather than membership; individuals may be covered by collective agreements when not union members. In 2001, the percentage affected by union pay agreements in Wales (41.6% for males and 41.5% for females) is higher than

25 The figures for the Objective One area were obtained from Nomis and are not comparable with the NES from ONS. 26 Source: Labour Force Survey.

20 England (32.8% and 34.9%), but similar to Scotland (41.6% and 45.5%). The decline in union membership has coincided with a fall in union militancy. In 1990, there were 85 days lost per 1,000 employees because of industrial stoppages, this fell to 6 days in 2000. These are almost insignificant compared to the early 1980’s when 2,903 days were lost per 1,000 employees.27

5. VACANCIES

Vacancy statistics are based on administration information supplied by job centres. The basic count used indicates the current number of unfilled vacancies that have been notified by employers. There are several reasons why regional vacancy statistics may be less reliable than other labour market indicators. The most serious drawback is that not all vacancies are notified to job centres but are advertised by different means. Comparisons between regions are distorted if the number of vacancies not notified to job centres varies by area. Another complication is that vacancies are allocated to job centres, rather than the region the vacancy occurs. For example, a vacancy arising in Port Talbot notified to a job centre in Bridgend would be classified as a vacancy in Bridgend not Port Talbot. The vacancy data were published by ONS until April 2001, but figures after this date are not available due to the introduction of Employer Direct by Jobcentre Plus, which transferred vacancy notification from local to regional level.

Table 11. Number of vacancies by country.

Vacancies Notified to Job Centres 1991-2000 1991 average 2000 average % Change Wales 10,640 13,070 +22.84 Scotland 21,396 25,527 +19.31 England 136,226 183,890 +34.99 GB 168,263 222,489 +32.23 Source: Nomis.

Between 1991 and 2001 there was an increase in the number of vacancies notified to job centres, this may be the result of increased vacancies and or changes in the recruitment patterns of employers. In Wales the number of vacancies increased by 20% over the period, less than the corresponding figures in England. The type of vacancy also differs between regions and the percentage of full-time vacancies in Wales is less than England or Scotland throughout the period. In July 2000, full-time vacancies represented 63.6% of all vacancies in Wales but 69.8% in England. Analysis of vacancies by industry (SIC92) shows that, when compared to England, the proportion of vacancies in manufacturing and particularly public administration, education and health is greater in Wales. In April 2001, the proportion of vacancies in public administration, education and health in Wales (29.6%) was double that of England (14.2%). A similar breakdown by occupation (SOC90) reveals a higher proportion of vacancies in personal/protective services in Wales (21.3% in October 2002), but a lower proportion of vacancies in distribution, hotels and restaurants and financial services. The structure of vacancies seems to indicate

27 Source: Office for National Statistics. Digest of Welsh Statistics Table: 7.12 Industrial stoppages at work, up to 1994 (a).

21 a continuation of the trends identified in the employment structure, a higher concentration in public sector at the expense of sectors such as banking and finance compared with the rest of the UK.

Figure 4.

Notified Vacancies in Wales, by Industry 1995-2001

Public administration,education & health t

n 80% e

m Transport and y o

l communications p m E

l 60% a Distribution, hotels and t o

T restaurants

h 40% S Construction

e g a t

n Manufacturing e c r e 20% P

Others

0% 1995 1996 1997 1998 1999 2000 Year

Source: Nomis.

Figure 5.

22 Notified Vacancies in Great Britain, by Industry 1995-2001

100% Other services s

e Public i c

n administration,education

c & health a V

Banking, finance and d e

i insurance, etc f i

o Transport and N

l communications a t o T

Distribution, hotels and

f 40% o

restaurants

e g a t n e

c 20% Construction r e P Manufacturing 0% 1995 1996 1997 1998 1999 2000 Year

Vacancies are distributed unevenly across Wales. WEFO (2001) notes that there are ‘hotspots’ notably in East Wales where filling vacancies appears to be a significant problem for certain employers. The localized nature of labour markets together with a degree of geographical immobility of labour, means that labour shortages could constrain the ability of certain areas to grow as fast would otherwise be possible.

6. UNEMPLOYMENT

General There are two commonly used measures of unemployment, which because of the different definitions of unemployment used, give different estimates of the number of people unemployed. The count of claimants of unemployment-related benefits records the number of people claiming unemployment-related benefits. The claimant count consists of all people claiming job seekers allowance or national insurance credits at employment service local offices. Such individuals must declare that they are out of work, capable of, available for and actively seeking work during the week in which their claim is made. According to the claimant count measure of unemployment; in September 2002 there were 46,400 people unemployed in Wales, representing 3.6% of the workforce.28 Claimant count unemployment has gradually fallen from over 10% at the trough of the last recession, to be lower in 2002 than at any other time in the last quarter of a century.

The International Labour Organization (ILO) measure of unemployment is the internationally agreed definition - the measure is used by the Statistical Office of the European Union and the Organization for Economic Co-operation and Development 28 Unemployment Counts and Rates, ONS Crown Copyright Reserved [from Nomis on 16 October 2002] – wholly unemployed claimants. Workforce base estimates.

23 (OECD) and other countries. The ILO measure classifies unemployed people as those individuals without a job, who want a job, have actively sought work in the last 4 weeks and are available to start work in the next two weeks or those who have found a job and are waiting to start it in the next two weeks. The ILO measure, provides an unemployment rate in Wales of 5.3%, which equates to over seventy-one thousand people. Unsurprisingly, movements in the two measures are highly correlated but as figure 6 illustrates the ILO measure has been almost uniformly higher the claimant count over the last decade and the gap between the two measures has grown from less than one percentage point in the early nineties to percentage points around the turn of the millennium.

A notable feature of the decline in unemployment throughout the UK in the 1990’s has been that unlike in previous decades there have been no significant increases in inflation as the labour market has tightened, suggesting that the natural rate of unemployment has fallen.29 Nickell (2001) attributes this fall to the decline in trade union power and changes to the benefit system, notably increasing strictness in determining eligibility and decline in generosity, measured relative to earnings.

Figure 6

Claimant Count and ILO Unemployment in Wales 1992-2002

10.0 ) % (

8.0 e t a R

e 6.0 m y o l p m

e 4.0 n U

0.0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Date

Claimant Count ILO Unemployment and Age

29 The natural rate of unemployment is the level of unemployment at which inflation stabilizes.

24 Typically three age categories are considered when looking at the age profile of unemployment - 16-24, 25-44 and over 45. The proportion of the unemployed who are aged between 16-24 has fallen since 1991 throughout the UK, but by a smaller percentage in Wales. Relative to the UK, unemployment in Wales has become more concentrated in the 16-24 age group - 31.4% in September 2002, compared to 27.4% in England.30 In contrast, the proportion aged between 25-44 has fallen in Wales, whilst remaining fairly stable for Britain as a whole. However, in Wales this proportion (44.3% in September 2002) remains below England (48%) and Scotland (46.4%). The proportion of the unemployed aged 45 or over in Wales rose from 17.6% in January 1991 to 24.7% in January 2002. The increase in Wales has been greater than in the rest of the UK, contributing to convergence in the proportion unemployed in the eldest group among countries within the UK.

Unemployment and Gender Throughout the nineties, the male unemployment rate exceeded its female counterpart, a feature common to other UK regions. For example, in Wales the male unemployment rate was 10.6% in January 1991, more than double the female rate of 4.0%. Similar to other UK regions, total unemployment has halved over the period, but female unemployment has experienced a slightly greater decline than male (falling to 6% for males and 1.9% for females respectively in January 2002). Thus, unemployment gender differentials have increased slightly in Wales, similar to England and Scotland. However, the gender difference in unemployment rates is greater in Wales throughout the period (the ratio of male to female unemployment was 3.16 in Wales, 2.65 in England and 2.95 in Scotland in January 2002).

Figure 7

Gender Differences in Claimant Count Unemployment in the UK 1992-2002

12 e t a

t n u o 8 C

m 6 i a l C 4

0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Year

England-Males Scotland-Males Wales-Males England-Females Scotland-Females Wales-Females Part of the explanation for the unemployment differential may lie with structural changes mentioned earlier, notably the decline in manufacturing industry, traditionally associated

30 Using the claimant count measure.

25 with male employment, and the rise in employment in the services sector, where females have greater representation.

Unemployment Duration The problem of long-term unemployment has received a great deal of attention from economists and policymakers because of the costs associated with having individuals without work for long periods. Unemployed workers are not producing output and this represents a waste of scarce resources. Okun's law states that for every 1% by which unemployment exceeds the natural rate, a GDP gap31 of 2% appears. Paying welfare benefits is a burden on taxpayers. Bivand (2001) estimates that the cost to the Treasury (and therefore ultimately the tax-payer) is over £10,000 per unemployed person per year.

Long-term unemployment may be self-perpetuating because human capital might depreciate whilst the worker is unemployed (see Roed 1997 for an overview). Authors such as Heckman and Borjas (1980), Blanchard and Diamond (1994) and Arulampalam, Booth and Taylor (2000), have all commented on the negative impact long-term unemployment has on re-employment prospects, whilst Gregory and Jukes (2001) investigate the wage penalty on individuals who do become re-employed. Moreover, Pissarides (1992) formulates a model where the depreciation of human capital lowers the average quality of unemployed workers and results in firms reducing the number of jobs created. Thus, short-run variations in labour demand can have permanent effects, which is encapsulated in the hysteresis paradigm (see Cross 1988 for a summary). Hysteresis can work through a number of alternative channels such as via an insider-outsider process, mismatch between workers and job opportunities and through low levels of capital formation.

Of the forty six thousand people unemployed in Wales in September 2002, 30% (14,247) had been without work for over six months and just under half of these (7,014) had been unemployed for over a year. These proportions are broadly similar to those for the rest of the UK. However, they have changed dramatically over time. In the ten years up to the summer of 1997, typically half of the unemployed had been without work for over six months and half of these again were unemployed for over a year, the actual proportions varying counter cyclically. The proportion of the workforce who are long-term unemployed32 has been declining throughout the UK since the last recession in 1992. Immediately after that recession and up to end of 1996, the proportion of long-term unemployed declined faster in Wales than in the rest of the UK. From 1997 onwards although the proportion continued to fall, it did so more slowly so that the rest of the regions in the UK were able to ‘catch-up’, giving the similarity of long-term unemployment rates currently observed.

It should be noted that extended spells of unemployment might have beneficial effects. It may also be that individuals voluntarily choose jobs that are characterised by frequent layoffs but have an associated compensating wage differential (Feldstein, 1978). Another benefit, initially proposed Alchian (1969), is that extended periods of job search lead to

31 A GDP gap is the difference be actual and potential GDP. 32 Those unemployed for over 12 months.

26 more efficient matching so that workers are able to find jobs in which they are more productive and receive higher wages (Altonji and Sakato, 1987; Mortensen, 1986 and Topel 1991). Tobin (1972) rejects this idea, arguing that employed workers get better information about jobs than those who are unemployed. Evidence from the UK does suggest that individuals who spend more time unemployed and searching do find a better match in their subsequent employment (Boheim and Taylor, 2000). However, they also find that job tenures following a spell of unemployment are, on average, much shorter than where the individual has entered directly from another job. Individuals who enter a job from unemployment are four times as likely to re-enter unemployment than those who switch voluntarily from another job.

Comparisons With Rest of UK Except for brief periods in 1993 and 1994, the unemployment rate in Wales has been higher than that in the rest of the UK, especially England, for the last two decades. This is consistent with the analyses of Brown and Sessions (1997) who find, after controlling for many demographic characteristics, there are significant regional disparities in unemployment risk, with individuals in Wales, Northern Ireland, the North and West Midlands of England facing a higher risk of unemployment, ceteris paribus.

In the 1990s the gap narrowed considerably, reflecting a convergence in unemployment rates across regions in the UK, continuing the trend identified after the recession of the early 1990s (Evans and McCormick, 1994). For example, in late 2002, unemployment rates in the UK ranged from 5% in the North East to 1.7% in the South East. In the late 1980s the differential between these two regions was over eight percentage points though with higher absolute levels of unemployment. At present, only the South East, Eastern England and the East Midlands have lower claimant count unemployment rates than Wales. The unemployment rate in Wales has tended to be similar to that in Scotland, with the difference between the claimant count rates in the two countries not exceeding one percentage point over the last decade.

Intra-regional Variations in Unemployment The convergence of regional unemployment has raised the significance of intra-regional differences. The unemployment claimant count is reported by Unitary Authority areas within Wales for males, females and overall. The annual averages, in table 12, are formed using monthly data (rates are not available until 1996).

Unemployment in each unitary authority exhibits a downward trend from 1996, as identified for Wales as a whole. However, the percentage fall in unemployment differs between areas and substantial intra-regional differences remain. Unemployment is prevalent in three main areas, the former mining and steel making areas in the South Wales valleys, rural West Wales and North West Wales. The restructuring of industry has caused structural unemployment in many areas, most notably the Welsh Valleys, where sectors employing large proportions of the area’s workforce have declined rapidly. In contrast, urban centres such as Cardiff, commuter areas such as the Vale of Glamorgan and areas neighbouring England (e.g. Flintshire, Wrexham, Powys and Monmouthshire) have lower unemployment rates. In 1996, for example, the male unemployment rate

27 ranged between a maximum of 17% in Blaenau Gwent and 5.6% in Powys, whereas in 2002 the monthly averages are 10.2% and 2.6% for the same areas, indicating also persistence in the relative position of areas.33 Indeed, Blaenau Gwent, the area with the highest unemployment rate, exhibits the smallest percentage decrease in unemployment from 1991 to 2002, for both males and females. Intra-regional dispersion is also evident for females, but to a lesser extent. Anglesey has the highest female claimant count rate 7% in 1996 and 3.6% in 2002. In 1996, Torfaen had the lowest unemployment at 2.8%, but by 2002 the relative positions changed, with Monmouthshire having the lowest claimant count rate of 1.1%.

The figures confirm a continuation of an East/West divide identified in the 1980s by Morris and Wilkinson (1989, 1993) and more recently discussed, with respect to North Wales, by Minford and Stoney (2001). In 2002, unemployment rates in Angelsey, Gwynedd, Pembrokeshire and Carmarthen all exceed the Welsh national average, with Flintshire, Wrexham, Powys and Monmouthshire below it. This is reflected in disparities in GDP within Wales. For example, in 1998 GDP per head is 70% of the UK average in West Wales, in contrast to 98% in east Wales. Location, particularly distance from and access to large urban markets and infra-structural endowments, seems to be important in determining the investment decision of companies and thus economic success of areas. Morris (1995) argues growth in South Wales, particularly in Cardiff, has been concentrated along the M4, which has attracted significant inward investment. In North Wales, growth again stems from the East (along the A55) resulting in lower unemployment and higher GDP in these areas.

Table 12. Unemployment by Unitary Authority34 Total unemployment claimant count year averages for unitary authorities in Wales % number change % rate change 1991 1996 2002 1991-2002 1996-2002 Number Number % Number % Anglesey 3326 3182 12.5 1629 6.6 -51.02 -47.2 Blaenau Gwent 3137 2889 12.7 1768 7.2 -43.64 -43.31 Bridgend 5466 3775 6.8 1887 3.5 -65.48 -48.53 Caerphilly 7180 6288 11.3 3017 5.1 -57.98 -54.87 Cardiff 13761 12756 7.0 5522 2.9 -59.87 -58.57 Carmarthenshire 5700 5459 7.1 2651 4.5 -53.49 -36.62 Ceredigion 1887 2011 5.9 900 2.7 -52.31 -54.24 Conwy 3467 3929 8.2 1684 3.8 -51.43 -53.66 Denbighshire 2971 3217 7.5 1230 2.8 -58.60 -62.67 Flintshire 3938 3884 5.9 1861 2.6 -52.74 -55.93 Gwynedd 5021 5383 9.3 2413 4.6 -51.94 -50.54 Merthyr Tydfil 2895 2521 11.0 1217 5.8 -57.96 -47.27 Monmouthshire 2229 2124 5.7 885 2.2 -60.3 -61.4 Neath Port Talbot 4857 4280 7.7 2408 4.9 -50.42 -36.36 Newport 6440 5762 8.5 2936 3.6 -54.41 -57.65 Pembrokeshire 4960 5083 10.9 2290 5.3 -53.83 -51.38 Powys 2919 2665 4.4 1380 2.1 -52.72 -52.27

33 Using the claimant count. 34 The rate change could only be calculated from 1996.

28 Rhondda, Cynon, Taff 10722 8331 9.5 3618 4.3 -66.26 -54.74 Swansea 9660 8089 8.1 4066 3.9 -57.91 -51.85 Torfaen 4000 3071 6.0 1509 3.7 -62.28 -38.33 Vale of Glamorgan 4279 4120 8.2 1930 3.8 -54.9 -53.66 Wrexham 4355 3848 6.5 1696 2.8 -61.06 -56.92

The housing market has also been found to contribute to regional unemployment disparities because of its influence on labour mobility. Classical economic theory postulates that if labour is perfectly mobile then differences in unemployment rates between regions should result in workers in the high unemployment areas looking for work in the low unemployment areas. Thus, with perfect labour mobility regional disparities in unemployment rates should only be a temporary phenomenon. However, council housing and other forms of social housing can impede the geographical mobility of labour by ‘locking’ council house tenants into the labour market in their immediate travel-to-work area. This can be accentuated by subsidised rents in the social housing sector that make council house tenants reluctant to move to find work in another locality (Hughes and McCormick 1981). In 2001, almost one-fifth of the housing stock in Wales was rented from local authorities, new town corporations or registered social landlords, similar to the proportion in 1991 though very different from 1981 when the figure was almost 30%.35 Higher concentrations of social housing are associated with high unemployment areas. For example, in Blaenau Gwent 30% of dwellings are socially owned. Similarly in Torfaen the figure is 28% and in Merthyr Tydfil, 26%. In contrast, low unemployment areas such as the Vale of Glamorgan, Ceredigion and Monmouthshire have low concentrations of social housing (13%, 11% and 15% respectively). The growth of owner occupation and associated increased transaction costs when moving limit search location and increase commuting distances (Oswald 1996). Migration from some parts of Wales to other parts of the Principality and the rest of the UK is hindered by house price differentials. For example, between April and June 2002 the average price of a house in Cardiff was £111,083, compared with £41,726 in Blaenau Gwent.36

The monetary and personal costs associated with commuting mean the majority of people work within the local authority where they reside. For example in Wales in 2001, 73% of all working residents in Wales worked and were resident in the same local authority. However, Cardiff had the largest net commuting inflow (42,000) with the majority of the corresponding gross inflows coming from the surrounding areas of the Vale of Glamorgan (19,000), Rhondda, Cynon, Taff (17,000) and Caerphilly (12,000). Regions close to the border with England also have large net outflows of workers with 18% of workers from Monmouthshire working outside of Wales and 26% in Flintshire. In most of the areas with Objective One status there are net outflows of workers.

Poor access to transport can limit the area of job search. The Welsh Assembly 37 finds the mode of transportation for travel to work depends on which area in Wales is considered. Those living in rural areas have less access to public transport. For example, in Powys

35 Source Welsh Housing Statistics 2001 table 1.3. 36 Data from Proviser http://www.proviser.com/property_prices/ 37 Welsh Assembly, Single Programming Document, East Wales Objective Two and Transitional Programme. Source 1991 Census.

29 only 2.6% use public transport to get to work compared to 16.5% Cardiff and 14.6% in Newport. Compared to England, people in Wales as a whole are more heavily dependent on car transport to get to work, with 77.9% of people in Wales travelling to work by car, in 2001, compared to 69.3% in England, where greater use is made of both buses and trains.38 Access to a car is higher in low unemployment regions. For example, in 1991, 43% of households had no car in Merthyr Tydfil and Blaenau Gwent, compared to 21% in Monmouthshire.39 As with many other social problems associated with unemployment, there is a question over the direction of causation. For example, are workers unemployed because they lack access to cars or do they not have access to a car because they are unemployed?

West Wales and the Valleys has qualified for Objective One funding between 2000-2006 because GDP in these areas is currently less than 75% of the EU average. The additional funds aim to reduce the disparities in GDP by increasing economic activity and employment. If correctly used, these funds also provide an opportunity to reduce the East-West divide in Wales. The remaining areas in Wales have qualified for Objective Three funding, which aims to increase education, training and employment. The Objective One area has higher unemployment40. In January 1996 the unemployment rate for the Objective One area was 9.3%, compared with 7.3% in non-Objective One area. Since then the rates have fallen for both groups, so that by January 2002 the respective figures were 4.8% and 3.0%. Unemployment fell proportionally more in the non- Objective One (Objective Three) area until 2000, since when the Objective One area has experienced a faster decrease in the unemployment rate. These results are in line with the improvement in economic activity in the Objective One area after 2000. Since 2000, the Objective One are has outperformed the Objective Three area in that unemployment and inactivity have fallen more quickly, a reversal of trends in the end of the last century.

Intra-Regional Differences in Unemployment Duration For all areas over the sample period the proportion of male long-term unemployed exceeds that of females. The proportion classified as long-term unemployed41 also exhibits geographical variations. For males, Merthyr Tydfil had the highest proportion of long term unemployed in 1991, with 1 in 3 of unemployed men being unemployed for over 12 months. Powys exhibits the lowest proportion in long-term unemployment (14.4%). By 2002, the situation has changed considerably; Anglesey has the largest proportion (38.7%) and Rhondda, Cynon, Taff the lowest (9.8%). For females, the long- term unemployment rate was highest in Anglesey in both 1991 and 2002. In 1991 Powys had the lowest proportion of females in long-term unemployment, but in 2002 this changed to Wrexham. Unlike unemployment, which fell over the period for all areas, the proportion of males in long-term unemployment rose in seven, predominantly rural areas, where long-term unemployment, although reduced since the early recession, had not fallen to 1991 levels. In contrast, areas such as Rhondda, Cynon, Taff, Caerphilly, Cardiff and Merthyr Tydfil experienced substantial falls in the proportion of long term unemployed (between 45-65%). Similar results are evident for females. 38 QLFS, Autumn 2001. 39 Mapping Social Exclusion in Wales, The National Assembly of Wales Statistical Directorate. 40 See also the differences in economic activity between Welsh Objective One and three areas. 41 Those unemployed for over a year.

30 The most significant change for all areas over the period is the rise in the proportion of unemployed aged 45 or above. Significant intra-regional differences emerge with particularly high unemployment growth rates for males in the previously heavy industrialised areas of Merthyr Tydfil (71.03%) and Neath and Port Talbot (63.57%). However, for females, the greatest rise is experienced in the rural areas of Anglesey, Monmouthshire and Ceredigion. It is, indeed, these rural areas that have the largest proportions of unemployed aged 45 or above, consistent with the decline of agriculture, combined with geographical and occupational immobility of labour.

Traditionally, policy makers have been concerned with high levels of unemployment amongst young people, because the usual costs of unemployment are accentuated by the detrimental impact on future labour market prospects (Gregg, 2001). There has been a 67% fall in youth Job Seeker Allowance claimants since the launch of the New Deal programmes in Wales. Figures presented in Hansard state that 24,800 young Welsh people have secured jobs through the New Deal since its launch and 80 per cent of those have retained their jobs for 13 weeks or more.42 The New Deal may also have had beneficial effects for other groups. For example, between July 1998 and March 2002, 24,700 long-term unemployed people joined the New Deal scheme. Of these 2,400 were still participating at the end of June 2002, whilst 6,300 secured jobs, either on or when leaving, any stage of the New Deal process. This included 5,300 (84%) who had entered jobs that lasted for at least 13 weeks. In October 1998, the New Deal was extended to lone parents, and by June 2002 the uptake reached 23,700. Half of these (11,900) had gained employment since joining the programme. Of course, those who found jobs after participating in the New Deal may have found work anyway (the deadweight effect). Furthermore, the recruitment of New Deal participants may be at the expense of people employers would have recruited anyway (substitution effect) or may be at the expense of workers already in the firm (displacement). The effectiveness of the New Deal Scheme is evaluated at a UK level by Van Reenen (2001), who concludes that the scheme has been a modest success with the social benefits outweighing the social costs.

Falling unemployment and regional unemployment convergence suggest labour market conditions have improved over the decade. However, recently, some authors have criticised the definitions of unemployment arguing that joblessness remains, but has become ‘hidden’ undetected by traditional measures (Beatty et al.2000). The fall in male unemployment has coincided with an increase in male inactivity, particularly among those classified as long-term sick. Wales contains both former coalfields and disadvantaged rural areas, where UK studies have found unemployment has been reduced as a consequence of rising inactivity (Beatty and Fothergill, 1996 and Beatty and Fothergill, 1997). Beatty et al. (2002) suggest hidden unemployment is particularly severe in the Welsh Valleys, where the ‘real’ unemployment rate is estimated to exceed 20 per cent, (Merthyr Tydfil has the highest level of ‘real’ unemployment at 28.2%), compared with low figures, of about 3%, in South East England.43 Indeed, for Wales as a

42 (Hansard 17 Jul 2002 : Column 319W) 43 The real unemployment rate is estimated by including those on government schemes, early retirees and sickness claimants, compared to fully-employed benchmarks in the South East.

31 whole the ‘real’ unemployment rate (13.3% in January 2002) is more than double the claimant count rate for the same period (5.6%).

Other commentators have highlighted the emergence of a grey area around the traditional labour market, in which workers are not fully integrated into employment and yet are not economically inactive; for example, government training schemes (Peck, 1990), informal work (Gershuny, 1979), home working, hidden or underground work (Finnegan, 1985) or work which is not recognised or unpaid (Morris, 1988). In this view, unemployment emerges as the result of sociological pressures rather than simple supply and demand interactions. However, measuring the ‘black economy’ is notoriously difficult and thus no comment can be made here about regional differences.

7. SUMMARY AND CONCLUSIONS

The 1990’s have witnessed an improvement in labour market performance in Wales and the rest of the UK, driven by an upswing in the wider economy. However, a number of indicators suggest that Wales still performs badly compared to the rest of the UK. Regional convergence has been identified in unemployment rates. However, these measures have been criticised on the grounds that increases in inactivity and ‘hidden’ unemployment in Wales may disguise the real level of unemployment and that substantial regional differentials remain (Beatty et al.2002).

Several main areas of concern for Wales are highlighted in this report. Inactivity remains a serious problem in Wales, having shown no tendency to fall to UK levels. The concentration of inactivity in previously high unemployment areas suggests hysteresis effects may be important. The concentration of inactivity within certain groups, particularly the over 50’s, and spatial concentration may suggest barriers to activity specific to these groups.

O’Leary et al. (2002) find those who attain NVQ level 4 or equivalent are about four times less likely to be inactive compared to an individual with no qualifications. The lower average qualification level in Wales thus explains part of the regional inactivity gap. Wales has both a higher proportion with no qualifications (and indeed poor basic skills) and a lower proportion at NVQ 3 and 4.

In addition, theory suggests that activity levels depend on earnings. The recent trend in average earnings is a matter of particular concern. Average earnings in Wales have actually fallen relative to the UK, thereby increasing the regional wage gap. This may result from differences in employment structure, for example Wales has a far lower proportion employed in banking and finance, a high wage industry.44 However, earnings for the same job may differ across regions and more detailed investigation is warranted.

Equally important is the existence of disparities within Wales. The relative position of different areas in Wales shows remarkable consistency over the range of labour market

44 See Blackaby et al. (2001) for further earnings breakdowns. For example in 2000 the Wales-Great Britain wage differential is lower for manual than non-manual work.

32 indicators. High unemployment and inactivity coexist with low levels of education and earnings, contributing to socio-economic divisions within the Welsh economy. These indicators are clearly inter-linked; improvements in education may increase activity, employment and earnings; policy initiatives must take this into account.

The poor performance of some areas in Wales such as Merthyr Tydfil, which has the highest real unemployment rate in the UK (Beatty et al.2002), is a cause for concern. Attention has focused on industrial decline, particularly in the South Wales Valleys however further divisions are evident. In particular, an East West division has emerged, with Unitary Authorities in East Wales performing relatively better than those in the West. Further analysis of these labour market divisions, in particular the impact of Objective One funding is being undertaken by WELMERC. The key issue is the extent to which gaps in labour market performance, both within Wales and across the UK, are likely to change through policy initiatives.

Within this framework there is a need for further research into both supply and demand factors within the labour market. It is clear that an upgrading of the quality of human capital will be necessary to increase employability but less is known about the precise distribution of increased skill requirements. Given the diversity within different regions of Wales and across Unitary Authorities the scope for disaggregated analysis allowed by the Welsh boosters to the British Household Panel and Labour Force Survey will be an important element in the future analysis of the Welsh Labour Market.

33 APPENDIX 1: DESCRIPTION OF THE DATASETS USED.

The Labour Force Survey (LFS)

The Labour Force Survey (LFS) is a quarterly sample survey of households living at private addresses in Great Britain carried out by the ONS. The LFS is based on a systematic random sample design. Each quarter, a sample of 60,000 private households (approximately 120,000 people over the age of 16) is made up of 5 groups. Each group is interviewed for 5 consecutive quarters. Thus, in any quarter, one group will have their first interview and one their last, resulting in an overlap of 80%. The results from the LFS are weighted, which corrects for those groups that are under-represented in the samples. Survey estimates are also produced, which reflect the population as a whole. Results are released for quarters; March to May, June to August, September to November and December to February.

The two main strengths of these data are the large sample size and frequency of collection. In addition, definitions meet international standards and are therefore comparable across countries. Results are available at the unitary authority level. However, at this spatial level, the sample sizes can be small and, therefore, the estimates become less precise. To reflect this, LFS estimates are rounded to the nearest thousand and sample sizes based on an estimate of less than 10,000 are suppressed. Many indicators are, therefore, unreliable at the unitary authority level within Wales. However, from April 2001 the Welsh Assembly has sponsored an increase in the Welsh sample, from 4,900 households to 18,000 households per year. The local Labour Force Survey is aggregated and produced annually. The most recent survey (2001/2002) includes the Welsh booster sample.

The New Earnings Survey (NES)

The main purpose of the survey is to obtain annual information about the levels and distribution of earnings of employees in various industries and occupations. A large sample of approximately 170,000 is used. The employer completes the questionnaire, and this is thought to increase accuracy relative to individual reports. The make up of pay is classified, using the following headings: overtime, profit related pay, payment-by- results, premium pay (e.g. night or weekend work) and basic pay. The sample excludes individuals who are not on PAYE schemes, in particular those below the income tax level, for example part-time workers. The data sets utilised in this study are from the New Earnings Survey - Analyses by region, county and small areas, as this provides the most comprehensive and detailed information at the sub-regional level.

34 APPENDIX 2: GLOSSARY

Claimant count

The claimant count consists of all people claiming job seekers allowance or national insurance credits at employment service local offices. They must have declared that they are out of work but seeking employment. The claimant count stock is used here, which provides the number of claimants on a single day each month. The unemployment rates are calculated using standard base rates, which are the sum of the workforce jobs and the claimant count number. Employee jobs, self-employment jobs, armed forces and government-supported trainees are included in workforce jobs.

ILO unemployment

ILO classifies people in three states, in employment, ILO unemployed or economically inactive. The ILO unemployed want to work but are currently without employment but have sought work in the last 4 weeks and can start work in the next 2 weeks or who are out of work but have gained employment and will start in the next 2 weeks. The ILO unemployment rate is the proportion of the economically active, those employed or ILO unemployed, who are unemployed.

Geographical areas

Countries and standard regions

Data are collected at a wide range of spatial scales, the two alternative sub-regional area definitions used in this report are, unitary authorities, of which there are 22 in Wales and travel-to-work areas (TTWA), which, using the 1998 definition, total 36 in Wales. Travel-to-Work Areas (TTWAs) are approximations to self-contained labour markets (at least 60% of those who live in the area work in the area) based on Census 1991 commuting to work patterns.

35 REFERENCES

Alchian, A.A (1969), “Information Costs, Pricing and Resource Utilization”, Western Economic Journal, 7 (2): 109-128.

Altonji, J.G. and Sakoto, R.A. (1987), “Do Wages Rise With Job Seniority?”, Review of Economic Studies, 54: 437-459.

Arulampalam, W., Booth, A.L. and Taylor, M.P. (2000), “Unemployment Persistence” , Oxford Economic Papers, 52: 24-50

Baran, D. and O’Donoghue, J. (2002) “Price Levels in 2000 for London and the Regions Compared with the National Average”, Economic Trends, No 578.

Basic Skills Agency (2001), “The Extent of the Problem: Basic Skills in Wales.”

Bean, C.R. (1994) “Unemployment in Europe: a Retrospective”, European Economic Review, 38(3): 523-534.

Beatty, C. and Fothergill, S. (1996) “Labour Market Adjustment in Areas of Chronic Industrial Decline: the Case of the UK Coalfields”, Regional Studies, 30, 627-640.

Beatty, C. and Fothergill, S. (1997) “Unemployment and the Labour Market in Rural Development Areas”, Rural Research Series, No 30. Rural Development Commisson, London.

Beatty, C. Fothergill, S. and Macmillan, R. (2000) “A Theory of Employment, Unemployment and Sickness”, Regional Studies, 34(7): 617-630.

Beatty, C. and Fothergill, S. (2002) “Hidden Unemployment Among Men: A Case Study”, Regional Studies, 36(8): 811-823.

Beatty, C., Fothergill, S., Gore, T., and Green, A. (2002) “The Real Level of Unemployment”, Centre for Regional Economic and Social Research, Sheffield Hallam University.

Bivand, P. (2001), “The Cost of Unemployment”, Unemployment Unit Working Brief 120, January 2001.

Boheim, R. and Taylor, M.P. (2002) “The Search for Success: Do the Unemployed find Stable Employment?”, Labour Economics, 9(6): 699-820.

Blackaby, D.H., Bladen-Hovell, R.C. and Symons, E. (1991) “ Unemployment, Duration and Wage Determination in the UK: Evidence from the FES, 1980-86”, Oxford Bulletin of Economics and Statistics, 53(4): 377-399.

36 Blackaby, D.H. and Hunt, L.C. (1992) “The ‘Wage Curve’ and Long Term Unemployment: a Cautionary Note”, Manchester School of Economics and Social Studies, 60(4): 419-428

Blackaby, D., Charles, N., Davies, C., Murphy, P., O’Leary, N. and Ransome, P. (1999) ‘Women in Senior Management in Wales’, Research Discussion Series, Equal Opportunities Commission, 1999, Manchester, pp.130.

Blackaby, D., Moore, N., Murphy, P., and O’Leary, N. (2001) “The Gender Pay Gap in Wales”, EOC Research Discussion Series.

Blackaby, D., Murphy P. D., and O’Leary N. C. (1999) “The Payment of Public Sector Workers in the UK: Reconciliation with North American Findings”, Economics Letters, 65(2): 239-243.

Blanchard, O.J. and Diamond, P. (1994), “Ranking, Unemployment Duration and Wages”, Review of Economic Studies, 61(3): 417-434.

Brooksbank, D. (2002) “The Welsh Economy: A Statistical Profile”, Contemporary Wales, 14: 164-192.

Brown, S. and Sessions, J. G. (1997) “A Profile of UK Unemployment: Regional versus Demographic Influences”, Regional Studies, 31(4): 351-66.

Cameron, G., Muellbauer, J., and Snicker, J. (2002), “A Study in Structural Change: Relative Earnings in Wales Since the 1970’s”, Regional Studies 36(1): 1-11.

Cross, R. (1988) Unemployment, Hysteresis and the Natural Rate Hypothesis (Blackwell, 1988)

Daniel, P. (1990), The Unemployed Flow, London: Policy Studies Institute.

Devine, T. and Keifer, N. (1991), Empirical Labour Economics: The Search Approach. New York: Oxford University Press.

Dickey, H. (2001) “Regional Earnings Inequality in Great Britain: A Decomposition Analysis”, Regional Studies, 35(7): 605-612.

Evans, P. and McCormick, B. (1994) “The new pattern of regional unemployment: causes and policy significance”, Economic Journal, 104: 633-647.

37 Felstead, A., Gallie, D. and Green, F. (2002) Work Skills in Britain 1986-2001, London, DFES and SKOPE,

Finnegan, R. (1985) “Working outside formal employment” in Deem, R. and Salaman, G. (eds) Work Culture and Society, Open University Press, Milton Keynes, 130-49.

Forsythe, F. P. (1995) “Male Joblessness and Job Search: Regional Perspectives in the UK, 1981-1993”, Regional Studies, 29 (5): 453-63.

Gregg, P. (2001) “The Impact of Youth Unemployment on Adult Unemployment in the NCDS”, Economic Journal, 111: F626-53

Gregory, M. and Jukes, R. (2001), “Unemployment and Subsequent Earnings: Estimating Scarring Among British Men 1984-94”, Economic Journal, 111: F607-625.

Gershuny, J. (1979), “The Informal Economy and its Role in Post- Industrial Society”, Futures, 10: 103-14.

Feldstein, M.S. (1978), “The Effect of Unemployment Insurance on Temporary Layoff Unemployment”, American Economic Review, 68: 834-46.

Fothergill, S. and Gudgin, G. (1982) Unequal Growth: Urban and Regional Employment Change in the United Kingdom, Heinemann Educational, London.

Fothergill, S. (2001) “The True Scale of the Regional Problem in the UK”, Regional Studies, 35(3): 241-46.

Haughton, G., Johnson, S., Murphy, L., and Thomas, K. (1993) Local Geographies of Unemployment, Avebury Press, Aldershot.

Haughton, G. and Lawless P. (1992), Policies for Potential: Recasting British Urban and Regional Policies, Regional Studies Association, London.

Haughton, G. and Peck, J. (1988) “Skills Audits: A Framework for Local Economic Development”, Local Economy, 3 (1): 11-19.

Heckman, J.J. and Borjas, G. (1980), “Does Unemployment Cause Future Unemployment? Definitions, Questions and Answers from a Continuous Time Models for Heterogeneity and State Dependence”, Economica, 47: 247-283.

Henley, A. (1998a), “Residential Mobility, Housing Equity and the Labour Market”, Economic Journal, 108: 414-427.

Henley, A., Disney, R. and Carruth, A. (1994), “Job Tenure and Asset Holdings”, Economic Journal, 104: 338-349.

38 Hildreth, A. (1999) “What Has Happened to the Union Wage Differential in Britain in the 1990’s?”, Oxford Bulletin of Economics and Statistics, 61(1): 5-32.

Hughes, G.A. and McCormick, B. (1981), “Do Council House Policies Reduce Migration Between Regions?”, Economic Journal, 91: 919-37.

Hughes, G.A. and McCormick, B. (1985), “Migration Intentions in the UK: Which Households Want to Migrate and Which Succeed?”, Economic Journal, 95:113-123.

Hughes, G.A. and McCormick, B. (1987), “Housing Markets, Unemployment and Labour Market Flexibility in the UK”, European Economic Review, 31: 615-641.

Manning, N. and Burdett, K. (1996) “Do Older Workers Drop Out of the Labour Market? Some Recent British Evidence”, Bulletin of Economic Research, 48 (2): 167- 171.

Midmore, P. (1999), “Rural Economic Development: A Research Agenda for the National Assembley for Wales”, Contemporary Wales, 11: 101-109

Midmore, P., Hughes, G. and Bateman, D. (1994) “Agriculture and the Rural Economy: Profiles Policies and Prospects”, Contemporary Wales, 6: 3-32.

Minford, P. and Stoney, P. (2001) “The UK and with it Cheshire and North Wales are on a Good Wicket but will Labours Re-Regulation Ruin it?”, North Wales and Cheshire Business Prospect, Liverpool Research Group in Macroeconomics 7(1).

Morris, J. and Wilkinson, B. (1989) “Divided Wales: Local Prosperity in the 1980s”, Report prepared for HTV Wales, Cardiff Business School.

Morris, J. and Wilkinson, B. (1993) “Poverty and Prosperity in Wales”, Report prepared for HTV Wales, Cardiff Business School.

Morris, L. (1988), “Employment, the Household and Social Networks” in Gallie, D (ed), Employment in Britain, Basil Blackwell, Oxford: 376-405.

Mortensen, D. (1986), “Job Search and Labor Market Analysis” in O.Ashenfelter and R.Layard eds. Handbook of Labor Economics, II, Ch 15. (Amsterdam: North Holland, 1986).

Munday, M. and Peel, M. (1997) “The Comparative Performance of Foreign-Owned and Domestic Manufacturing Firms During Recession: Some Descriptive Evidence From Wales”, Contemporary Wales, 10: 50-79.

Nickell S. (2001) “Has UK Labour Market Performance Changed?”, Speech to the Society of Business Economists 16 May 2001.

39 National Assembly for Wales, ‘National Economic Development Strategy’ (2002)

O’Leary, N., Murphy, P. Moore, N., Manning, N., Jones, M., Harris, S., and Blackaby, D. (2002) “Inactivity and Unemployment in Wales and the Prosperity Gap”, Report for the WDA.

Oswald, A. (1996), “A Conjecture on the Explanation for High Unemployment in the industrialised Nations: Part 1”, Warwick Economic Research Paper, no. 475

Peck, J. (1990), “The Youth Training Scheme: Regional Policy in Reverse?” Policy and Politics, 18: 135-43.

Pissarides, C.A. (1992), “Loss of Skill during Unemployment and Persistence of Employment Shocks”, Quarterly Journal of Economics, 107(4): 1371-1391.

Rees, G. and Thomas, M. (1991), “From Coalminers to Entrepreneurs? A Case Study in the Sociology of Re-industrialisation”, in Cross, M and Payne, G. (eds), Work and the Enterprise Culture, Falmer Press, London: 57-78.

Rice, P. (1999) “The Impact of Local Labour Markets on Investment in Further Education: Evidence from the England and Wales Youth Cohort Studies”, Journal of Political Economy, 12(2): 287-312

Roberts, A. (1994), “The Causes and Consequences of Inward Investment: The Welsh Experience”, Contemporary Wales, 6.

Robinson, H. (2001) “Change in the Shires? Regional Differences in the Gender Pay Gap in the Advent of the National Minimum Wage”, University of Wales, Cardiff.

Roed, K. (1997), “Hysteresis in Unemployment”, Journal of Economic Surveys, 11(4): 389-418.

Senior, M. L. (1998) ‘Area Variations in Self-Perceived Limiting Long Term Illness in Britain, 1991: Is the Welsh Experience Exceptional?, Regional Studies, 32(3): 265-280.

The Low Pay Commision (1998) ‘The National Minimum Wage’, First Report of the Low Pay Commission.

The National Assembly for Wales, “National Economic Development Strategy” Consultation Document.

Tobin, J. (1972), “Inflation and Unemployment”, American Economic Review, 62 (1):1- 18.

Topel, R.H. (1991), “Specific Capital Mobility and Wages: Wage Rises with Seniority”, Journal of Political Economy, 99: 145-176.

40 TUC (2001), ‘It’s your call’, TUC February.

Van Reenen, J. (2001) “No More Skivvy Schemes? Active Labour Market Policies and the British New Deal for the Young Unemployed in Context”, Institute for Fiscal Studies Working Paper W01/09.

Wales National Assembly and Basic Skills Agency (2001) “The National Basic Skills Strategy for Wales.”

WEFO (2001) “Equal: Final Wales Implementation Plan” The National Assembly for Wales, 30 April.