Summary

New Labour’s policies aimed at regenerating poor communities promote economic investment, yet research on how such investment improves employment prospects for local residents over a sustained period is scarce.  Using the Darnall area of Sheffield as a case study, this paper considers the degree to which new sectors of employment have been accessed by residents living in the immediate vicinity of a major retail development.  Evidence from the 1981, 1991 and 2001 Censuses includes population, employment, migration and travel to work characteristics, which show that many residents remain marginalised from local jobs, raising important issues for contemporary initiatives promoting retail investment as part of a strategy to tackle worklessness.

Three decades of regeneration strategies in the UK’s northern cities have sought to address the deprivation left in the aftermath of industrial restructuring. Major job losses in traditional manufacturing and production sectors and significant growth across service based industries, often influenced by global and technological developments, have involved new labour requirements (Payne and Payne, 1994; Martin and Morrison, 2003). This period of dynamic labour market change has resulted in marked distinctions between high and low quality jobs and particular challenges for those without the skills, qualifications, work experience and flexibility expected in many growth sectors (Gordon, 2003; Escott, 2009). In this context the impact of change has been uneven, with some social groups being marginalised and increasing numbers living in areas regarded as ‘uncompetitive communities’ (Gordon and Turok, 2005: 243). These neighbourhoods experiencing recurrent risks of high unemployment remain a particular concern for policy makers today (DCLG, 2009).

Locality specific research in communities seeking to readjust and recover from major structural change is limited, with the study of male employment changes in the UK’s coalfields being a rare example (Beatty, et al., 2007). This spatial dimension of job change with segregation and clustering of jobs in particular localities, is also lost in labour market analysis and analysis of regeneration approaches (Buckner, 2009). Within UK cities the concentration of unemployment in particular areas is usually associated with the education, qualifications and transferability of skills between sectors of employment (Gordon, 2003; SEU, 2004). Yet much of the employment policy focus centering on the question of skills and experience of local people overlooks the nature of the labour market itself, and the types of interventions required to address fundamental inequalities (Yeandle, et al., 2009).

One example of this failure is demonstrated by examining the assumptions behind retail-led regeneration initiatives. Although the benefits of economic investment in deprived areas have been shown to be limited, retail development is commonly promoted in regeneration schemes (DETR, 2000; North, et al., 2003; Claxton and Siora, 2008). The sector’s growth over three decades,1 making it the third largest service based industry in the UK (Dixon, 2005; Burt & Sparks, 2003), has stimulated the promotion of retail as a strong contributor to local economies (DTI, 2005; Emery, 2006). It is argued that the sector not only provides local jobs but can act as a catalyst for broader economic renewal (Dunford; 2006, Guy, 2008). The types of benefits expected in local communities usually include the objective of creating opportunities provided by retail schemes for residents who have traditionally depended on living in close proximity to their place of work. The apparently simple argument that locating jobs in deprived areas results in benefits for local residents remains unsubstantiated however and, whilst New Labour has shifted its regeneration focus to a more locality specific approach, the importance of retail in ‘place making’ remain important (Lowe, 2005; Parker and Garnell, 2006; DCLG, 2008).

The types of employment benefits retail is expected to bring for regeneration areas includes entry-level and part-time employment, often regarded as appropriate for those groups most likely to be unemployed including young people and those with dependents and caring responsibilities (GLA, 2005). And although there are numerous examples of schemes linking up unemployed local residents with retailers, evidence suggesting that the sector has brought employment benefits to deprived localities is limited (Carley, et al, 2001). In their study of retail and neighbourhood regeneration Carley et al (2001) argue that instead it often displaces local employment, suggesting that the measurement of the employment impact of retail development is a source of major uncertainty. The problem for many regeneration areas is that investment has not been sustained beyond initial recruitment and retail employers have not provided sufficient support structures to secure long term benefits (Escott, 2006). The argument that locality specific investment in retail creates jobs for local residents appears too narrowly defined and in order to understand local impact, a wider and longer term set of measures is needed.

Understanding neighbourhood changes

One way of analysing the impact of employment change on local communities is to move beyond an assessment of jobs created, to include wider social and locational issues for people living in areas adjacent to new developments. Evidence about the local impact of structural changes in labour markets (Peck, 1996; Gordon and Turok, 2005; Helms and Cumbers, 2006) and the influence of geography on access to work (Green and Owen, 2007) suggests that social and cultural differences, often exemplified in movement of employees within and between local labour markets, are important. This local dimension of labour markets, which has been found to be especially important for those living in deprived areas, involves understanding patterns of mobility, housing provision, and the location of workplaces. Ward level analysis allows for examination of this ‘lived experience’ of local labour markets which plays a crucial role in shaping communities (Raco, 2007). Along with location of schools and care services, these are facets of policy making which are often overlooked in discussion of the wider employment benefits of economic development and regeneration schemes (Fanstein and Servon, 2005; Greed, 2005). The research on NDC areas also suggests demographic changes where those with qualifications and jobs move out of deprived communities with ‘a tendency for those in employment to be replaced by those not in work’(Lawless, 2006: 1999). This paper aims to tease out these local changes by exploring labour market trends alongside social characteristics and transport patterns in a neighbourhood where there has been a net growth in jobs.

The research approach

Like many cities a major part of Sheffield’s economic readjustment to service based industries included retail sector growth.  But the question of how jobs created were distributed and accessed remains unclear. Sheffield’s experience in which predominantly male full-time manufacturing jobs were replaced by female dominated service sector jobs  is exemplified in the building of the Meadowhall shopping centre in the late 1980s (Work Foundation, 2007; Creative Sheffield, 2008). It also signified the beginnings of a regeneration approach designed to revive the communities in the east end of the city. Largely physical investment in infrastructure during this period improved access and transport in the Lower Don Valley, and paved the way for private sector developers to build Meadowhall on land vacated by former steel works. Analysis of the local resident profile in Darnall ward, the locality in which the investment was made, provides an opportunity to assess employment activity over a sustained period (1981-2001). In order to understand the locality, a number of related questions are also important.  Have the social characteristics of Darnall’s residents changed? Who works in Darnall and where do they come from? Have employment rates among local residents improved? What are the commuting patterns affecting the ward?

Methodology

The case study area

The selection of Darnall ward, the traditional heart of Sheffield’s steel industry, for this study can be justified from a number of perspectives. First, the demise of the UK’s steel sector in the 1980s had a very significant impact on local residents and since then the area has been associated with jobs in service and tertiary industries (Sheffield City Council, 2006a). Secondly, the primary focus of investment, Meadowhall shopping centre, built on the site of a former steel works on the northern edge of the ward is also easily accessible to the M1 motorway and to bus, rail and tram connections2 suggesting that the hinterland for labour supply for the ward’s employers is significantly wider now than prior to the most serious period of de-industrialisation. Thirdly, service jobs have replaced manufacturing jobs with the shopping centre alone employing 8,000 people, and together with other office, retail, leisure and call centre activities creating at least 20,000 jobs (Rimmington, 2005). In spite of significant private sector job growth and regeneration funding in the form of area-based and European initiatives, the ward’s economic activity rate at 68 per cent is below the city average of 72 per cent and Darnall remains one of Sheffield’s poorest communities, scoring zero per cent for ‘urban prosperity’ and 44 per cent of its population ‘live on moderate means’, much higher than the city average of 12 per cent (2001 Census Standard Tables, ONS, LASOS, 2009). It is also one of several wards in the city where training and quality of local jobs remain primary concerns for residents (Sheffield City Council, 2006a; Rimmington, 2004).

Figure 1: Wards surrounding Darnall in Sheffield and Rotherham

Source: 2001 Census Standard Ward boundaries. This work is based on data provided through EDINA UKBORDERS with the support of the ESRC and JISC and uses boundary material which is Copyright of the Crown.

Ideally for this study we would have constructed an area that included Meadowhall and a fixed radius around it. However, this would require using census data at the sub-ward level which are not geographically consistent over time and, in addition, the 2001 Census data were subject to small cell adjustment methodology which seriously restricts the usefulness of these data at very low geographical levels. This process was implemented as a measure of protection against the disclosure of information about individuals who might them be identified from this. However, this adjustment means that data at lower geographical levels are less likely to be accurate than data at higher geographies since the larger the geographical area of study, the less likely it is that small values will be recorded and subsequently adjusted (Stillwell and Duke-Williams, 2007).

In addition, Meadowhall cannot really be discussed in isolation since its development had a knock-on effect, leading to the opening of many small business in other parts of the ward which brought extra employment to the ward.

Data

In order to create a picture of the economic position of Darnall’s residents before, and after the opening of Meadowhall, standard output from the 1981, 1991 and 2001 Censuses has been accessed via NOMIS. In addition, Census Special Workplace Statistics (SWS) for the same years have been analysed. These include tables of migrant flows between origins and destinations available at three geographic levels: local authority districts; wards and sub-ward level. Here the ward-level flows are used to explore the characteristics of people who work in Darnall but who might not necessarily live in the ward, as well as showing where Darnall residents who do not work in the ward were commuting to. These data tables were accessed on-line through the Centre for Interaction Data Estimation and Research (CIDER) at the University of Leeds using their Web-based Interface to Census Interactive Data (WICID) (Stillwell and Duke-Williams, 2003).

Comparing census data over time can be problematic because of the changes to the population base, definitions of variables, banding of data and areal boundaries (Martin, et al., 2002). However, minor changes to Darnall’s boundary between the 1981 and 1991 Censuses have been taken into account using the SWS whenever possible since earlier data from this source can be accessed so that it corresponds to 2001 Census boundaries.

Results

Population profile 1981-2001

Darnall’s population (19,000 residents) appears relatively stable in that the area did not suffer from depopulation between 1981 and 2001 and in the year up to the 2001 census the ward had more in-migrants than out-migrants. Nevertheless, migration patterns resulted in significant changes in the age and ethinicity profile of the ward. Darnall has a high proportion of children and young people with over a quarter of residents (28 per cent) under the age of 16 in 2001 (up from 22 per cent in 1981) (Table 1). At the same time there was a decrease in people of working age (16-64) but little change in the population aged 65 or over. With 27 per cent from ethnic minority groups (2001 Census Standard Tables, ONS) Darnall is one of the city’s most culturally diverse wards with large Pakistani, Bangladeshi, Yemeni and Somali communities (Sheffield City Council, 2006b). The changing ethnicity pattern includes 14 per cent of Darnall’s residents who were born outside the UK in 2001. This aspect of the profile shows how over a period of two decades the characteristics of the local population have changed significantly, raising important issues about whether the regeneration approaches originally assumed to assist the ward remain relevant.

Table 1: Characteristics of the local population: Residents of Darnall ward, Sheffield (%)

	Darnall ward						Sheffield
	1981		1991		2001		1981		1991		2001
All residents
Age
0-15	22	20		28		21		18		24
16-64	61	62		56		62		63		60
65+	17	18		16		17		18		16
% Born outside the UK	9	9		14		4		4		6
% Non White(1)	–	17		27		–		5		9
% Resident in Social Housing	28	28		29		47		36		30
% No Access to a Car	45	41		31		43		35		27

Source: 1981, 1991 and 2001 Censuses, Crown Copyright.
Notes: (1) All BME groups, excluding White non British.
In the 1981 Census ethnic group data were not collected.

Regeneration strategies often associate social housing with high levels of unemployment and economic inactivity. In this context, it is interesting to note that the percentage of Darnall’s households living in social housing remained fairly constant during the period 1981-2001 at 28-29 per cent. Thus there has been no increase in social housing occupancy in the ward. This chimes with recent research which refutes the links made between social housing and worklessness, showing that labour marginality is caused by educational attainment, attitudes to work and labour market demands rather than by housing type (Hills, 2007; Fletcher, 2008).

In areas experiencing high levels of poverty, low car ownership and relatively high public transport costs can be an added constraint on labour market engagement (Daniel, 1990; Buckner, 2009; Escott and Buckner, 2006). Although Darnall’s residents are significantly more likely to be car owners than they were in 1981, almost a third did not have access to a car in 2001. However, most of the ward’s residents do live within walking or cycling distance of Meadowhall and can access a network of local buses. Thus physical access to local jobs does not emerge as a particular problem for Darnall’s residents, indicating that more fundamental social and economic barriers may be more important determinants of labour market inactivity.

Economic activity

Alongside changes in population characteristics, employment rates for different groups have altered over the period of study, reflecting some of the wider changes in the district and region. Whilst labour market activity among men living in Darnall remains higher than for women the profile (Table 2) shows a sharp decrease in employment amongst men, with a subsequent increase in economic inactivity, and in particular permanent sickness and disability. This pattern of male employment change was also seen in the former coalfields where ‘hidden unemployment’ is a significant challenge twenty years on from most of the pit closures (Beatty, et al., 2007). In spite of significant female dominated service sector growth there was remarkably little change for women living in Darnall, showing that female labour market participation did not replace male employment among local residents. Given the changes in ethnicity in the ward’s population, the lack of increase in women’s employment may be caused by a higher proportion of women who look after their home/family full-time.  But it may also be caused by problems accessing employment including lack of employer response, indirect discrimination and language skills, which have been found to be common for ethnic minority women (Yeandle, et al., 2006).

Table 2:  Economic activity of people aged 16+: Residents of Darnall ward, Sheffield

		1981		1991		2001
		Men	Women	Men	Women	Men	Women
Numbers	All residents 16+	7,171	7,568	7,018	7,666	7,318	7,831
	Economically active	5,488	3,234	4,909	3,451	4,599	3,407
	In employment	4,670	3,014	3,851	3,067	3,948	3,027
	Unemployed	751	207	954	299	504	238
	Permanently sick or disabled	186	86	458	319	588	355
	Other*	1,564	4,261	1,755	3,982	2,278	4,211
Percentages	All residents 16+	100	100	100	100	100	100
	Economically active	77	43	70	45	63	44
	In employment	65	40	55	40	54	39
	Unemployed	10	3	14	4	7	3
	Permanently sick or disabled	3	1	7	4	8	5
	Other*	22	56	25	52	31	54

Source: 1981, 1991 and 2001 Censuses, Crown Copyright.
Note: * For 2001, all people aged 75+ were assumed to be retired since the census did not record the economic activity status of people aged 75 or over.

Industrial change in Darnall has also affected the types of occupations and quality of jobs available in the locality. The high proportion of residents who were working in manufacturing (57 per cent of and 26 per cent of women (1981 Census SAS, ONS) reflects the heavy reliance on jobs in the steel industry, before the building of Meadowhall. By 1991, this had decreased to 28 per cent of men and 16 per cent of women and by 2001 had fallen again to 27 per cent of men and 10 per cent of women (1991 Census LBS, 2001 Census Standard Tables, ONS). By contrast, just five per cent of Darnall’s residents were in sales occupations (the fourth lowest percentage of Sheffield’s wards) in 1981. Twenty years on this proportion had doubled to over ten per cent reflecting the shifting employment base of the city as a whole but not the bigger shift in the proportion of local jobs available in the service sector.

The ward’s job growth, from 22,000 jobs in 1981 to 25,000 jobs in 2001, was however not reflected in increased employment among local residents. In fact the reverse occurred (Office for National Statistics 1981a, 1991a, 2001a). During the period in which Sheffield’s manufacturing base collapsed and Meadowhall was built (1981-1991), the proportion of Darnall’s residents employed in the ward decreased by over a quarter, from 35 per cent in 1981 (64 per cent in manufacturing) to 27 per cent in 1991 (37 per cent in manufacturing). Another decade later the proportion of residents working and living in Darnall had recovered a little (29 per cent) but did not reflect the substantial growth in service jobs located in the ward.

People commuting into Darnall 1981-2001

Given these changes, it is perhaps not surprising to find from the Special Workplace Statistics that the proportion of local residents who were employed in Darnall fell from 11.2 per cent of in 1981 to 8.6 per cent in 1991 (Office for National Statistics, 1981a, 1991a). The fact that this lower proportion remained the same during a period of significant employment growth (up to 2001) leads to the question of the characteristics of those who work in Darnall, including those who do not live in the ward.

Figure 2: People aged 16 to state pension age (64 for men and 59 for women) in employment in Darnall by age, sex and ward of residence: 1981, 2001

Source: 1981 Census: Special Workplace Statistics (Set C) (re-estimated for 2001 boundaries; incorporating pro-rate ‘workplace unstated’, 2001 Census: Special Workplace Statistics (Level 2).

Analysis of the age structure of people in employment in Darnall in 1981 and 2001 (Figure 2) shows that over this twenty-year period, there was a large decrease in male workers aged 25-64, and a parallel increase in women aged 25-59. This change was driven mainly by the change in the profile of people coming from outside Darnall to work in the ward.

Comparing the social class of people working in Darnall in 1981 and 1991, before and after Meadowhall was opened, shows that prior to it opening the majority of people commuting into the ward to work were men (81 per cent) who were mainly employed in skilled manual and partly-skilled jobs. By 1991, the proportion of men commuting into Darnall to work dropped to 67 per cent with a decreasing proportion in skilled manual and semi-skilled jobs partially offset by an increase in the proportion of men in managerial and technical jobs. The proportion of women people commuting into Darnall to work subsequently increased as a consequence of increasing employment in skilled non-manual and managerial and technical jobs (Figure 3). The analysis illustrates widening differences between high and low quality jobs which may affect local residents and their employment prospects.

Figure 3: People in employment in Darnall by social class, sex and ward of residence: 1981, 1991

Source: 1981 Census and 1991 Census: Special Workplace Statistics (Set C) (re-estimated for 2001 boundaries; incorporating pro-rate ‘workplace unstated’.

For people who work in Darnall and who also live in Darnall the same pattern is observed but to a much lesser degree, with a decrease in men’s employment in skilled manual jobs and an increase in women’ employment in skilled non-manual occupations. Since the majority of people working in Darnall are not local residents, the shift to more skilled employment has been to the advantage of people commuting from outside the ward.

Commuting patterns

Although the scale of commuting into Darnall to work has not altered significantly over the period the characteristics of in-commuters has, suggesting different patterns of access to the jobs available in the ward. The Special Workplace Statistics have been used to identify the changes over the period 1981-2001 (Office for National Statistics, 1981a, 1991a, 2001a). The method of commuting to work for people who live and work in Darnall has changed with a switch from walking or cycling mirrored by an increase in car use. A similar pattern is seen for commuters into Darnall but for these people the switch has been from public transport. This is at a time when public transport links into the ward improved with the opening of the transport hub at Meadowhall.

The wards that in-commuters live in have also changed. In 1981 most people who travelled into Darnall to work were resident in the surrounding wards (such as Manor and Intake). By 1991 this area had changed to include wards further afield such as Chapel Green (which has a rail link to Meadowhall) and Bramley, Ravenfield and Wickersley (which is on the motorway network which links to Meadowhall) at the expense of closer wards, in particular Manor, Intake and Brightside. However by 2001 the pattern had again changed with wards such as South Wortley rising in importance with the expansion of the Supertram network.

Table 3: Origin of Darnall’s workforce – 1981, 1991, 2001

Ward name	District		1981		1991		2001
			Number	% workplace population	Number	% workplace population	Number	% workplace population
Darnall		Sheffield	2,490	11.2	1,790	8.6	2,119	8.6
Mosborough		Sheffield	920	4.1	950	4.6	1,284	5.2
Handsworth		Sheffield	1,820	8.2	1,010	4.9	1,007	4.1
Brightside		Sheffield	740	3.3	660	3.2	801	3.2
Chapel Green		Sheffield	380	1.7	670	3.2	767	3.1
Brinsworth, Catcliffe and Treeton		Rotherham	890	4.0	870	4.2	751	3.0
Intake		Sheffield	1,120	5.0	650	3.1	708	2.9
Aston, Orgreave and Ulley		Rotherham	730	3.3	770	3.7	706	2.9
Birley		Sheffield	890	4.0	430	2.1	649	2.6
South Wortley		Sheffield	390	1.7	340	1.6	502	2.0
Manor		Sheffield	1,140	5.1	440	2.1	449	1.8
Kimberworth		Rotherham	430	1.9	500	2.4	421	1.7
Nether Shire		Sheffield	560	2.5	440	2.1	421	1.7
Thorpe Hesley		Rotherham	410	1.8	330	1.6	411	1.7

Source: 1981 Census and 1991 Census: Special Workplace Statistics (Set C) (re-estimated for 2001 boundaries; incorporating pro-rate ‘workplace unstated’, 2001 Census: Special Workplace Statistics (Level 2).

The limitations of local economic investment

In conclusion, the labour market for many of those people living nearest to one of the UK’s largest regional retail developments appears inaccessible. Recovery has not occurred as anticipated and twenty years on investment in retail and related service industries has had limited success in relation to improving labour market prospects for Darnall’s residents. Darnall experienced the greatest job growth of any ward in Sheffield over the period of study, yet the ward had the smallest decrease in people without work by some considerable margin. In Darnall the number of people not in employment (excluding students) fell just 2.3 per cent between 1981 and 2001 compared with other wards in Sheffield where economic inactivity reduced much more rapidly over the same period.

This neighbourhood-level study also shows that macro-level analysis used in many economic assessments to demonstrate job change hides many of the underlying issues. Examination of the spatial patterns of residents and workers in one locality has uncovered some of the indicators which lie behind persistently high levels of unemployment including migration patterns, higher levels of in-commuting, male skilled manual work being replaced with female skilled non-manual jobs, and new labour market opportunities taken by people commuting in from outside the ward. Physical access to local employment does not appear to be a particular issue for Darnall residents, suggesting other reasons for labour market marginalisation. Demographic changes have undoubtedly affected labour market engagement in Darnall, and some of the residents in 1981 and 1991 may have gained new jobs locally but then moved out of the ward, facilitated by good transport links and reflecting patterns found in the NDC areas (Lawless, 2006). This study has focused on labour demand and statistical data exploring the impact of industrial change in one ward. Further qualitative research examining labour supply and the experiences of those living in the area adjacent to Meadowhall is required in order to explore the underlying barriers to employment. The findings also suggest comparative approaches looking across with similar levels and types of investment are needed in order to assess whether the support available for local residents is effective and beneficial.

The complex demographic and economic processes operating in this community raise important questions about the effectiveness of the types of employment creation strategies still being promoted by policy makers. Failure to consider local investment in this broader sense seems likely to result in continuing problems for communities such as Darnall where the benefits of job creation are not forthcoming and local capacity to adjust to wider economic change, including employer engagement, remains problematic.