Accessibility and public service provision - Wiley Online Library

17 downloads 1700 Views 306KB Size Report
key words Wales public service provision accessibility post offices GIS. GIS Research ..... offer a traditional full-time service, and those tar- geted to become an ...
ransactions

of the Institute of British Geographers

Accessibility and public service provision: evaluating the impacts of the Post Office Network Change Programme in the UK Mitchel Langford and Gary Higgs The importance of public service provision and accessibility in shaping government policies aimed at enhancing social inclusion and ensuring social justice in the UK is well founded. The capabilities of GIS for generating information to address such concerns have facilitated a widespread interest in measuring and analysing accessibility to public services. Previous studies have drawn attention to the potential implications of any changes in the provision of essential services such as post offices, food shops or GP surgeries. For example, the number of post offices in the UK has declined significantly since the mid 1960s and GIS has been used to monitor spatial and temporal patterns in closures in previous research studies. The aim of this paper is to assess the implications of the 2007–08 Network Change Programme on the spatial configuration of post offices in Wales. Using a network analysis approach, we identify those areas that have no service within the access criteria used to guide the closure programme and calculate the extra distances involved in travelling to the nearest outlet following closures. Our results suggest that national guidelines on provision of post offices are not met in Wales or within any of the five area plans that cover Wales. Perceptions that it is rural areas that have been disproportionately impacted can be challenged on the basis of the measure used to examine such changes. Such findings can provide an important context to in-depth surveys of the impacts of the closure programme on potentially vulnerable groups within both urban and rural communities. Furthermore they demonstrate how GIS-based analysis can inform policymakers of the potential impact of changes in provision for those most dependent on a range of public services. key words Wales

public service provision

accessibility

post offices GIS

GIS Research Centre and Wales Institute of Social and Economic Research, Data and Methods, Faculty of Advanced Technology, University of Glamorgan, Pontypridd, Wales CF37 1DL email: [email protected]

revised manuscript received 1 April 2010

Introduction It is widely recognised that the concept of accessibility is complex and multi-faceted and that it includes aspects such as financial, socio-cultural and organisational barriers as well as spatial impediments (Gulliford and Morgan 2002). While there continues to be an on-going debate over exact definitions and the best methodology for calculating suitable accessibility metrics (see for example Talen 2003), it remains clear that geography and the relative spatial distribution of providers and consumers of services is a critical component. The

importance of accessibility to public services in shaping government policies aimed at enhancing social inclusion and ensuring social justice is well founded. For example, there is a requirement for strategies to ‘improve accessibility for all’ to be included as part of partnership approaches in the Local Transport Plans prepared by UK local authorities over the period 2006 to 2011 (Social Exclusion Unit 2003). Likewise the National Health Service Priorities and Planning Framework 2003–2006 included a requirement to ensure equal access to primary and secondary health care services for disadvantaged groups (Department of Health 2002).

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

586

Further recognition of the importance of accessibility is evidenced by the inclusion of a ‘Geographical Access to Services Domain’ within each of the multiple deprivation indices currently being used by the various administrations of the UK and other organisations in targeting resources to particular areas (Noble et al. 2006). This component is currently based on drive-time proximity to various key services, such as post offices, general food stores, GP surgeries and primary schools. The analytical capabilities of GIS for generating information such as drive-time proximity have led to a widespread interest in measuring and analysing accessibility to a range of essential services. Such studies are often conducted within the context of providing evidence of social exclusion or for revealing patterns of spatial inequity with respect to areas with significant numbers of potentially vulnerable people. For example, recent studies have used spatial analytical approaches to investigate accessibility to health care facilities (Wang and Luo 2005; Yang et al. 2006; Martin et al. 2008), to public transit opportunities (Currie et al. 2009; Lei and Church 2010), to affordable food sources (Larsen and Gilliland 2008; McEntee and Agyeman 2010), to essential financial services (Leyshon et al. 2008), to health-promoting land use activities (Omer 2006; Comber et al. 2008; Jones et al. 2009) and to a variety of other public services (Halden 2002; Witten et al. 2003; Apparicio and Seguin 2006; Langford et al. 2008). A Department for Environment, Food and Rural Affairs (2004) report highlighted the problems facing those groups who need to access key public services, particularly in rural areas, and drew attention to the potential implications of any changes in the provision of services such as post offices, food shops or GP surgeries in relation to a range of internal and external forces. The importance of the post office in particular has been widely recognised since it is often one of the few, if not the only, service that remains in many rural villages. In May 2007 Post Office Ltd announced the implementation of the Network Change Programme aimed at reshaping the UK post office branch network (Department for Business, Enterprise and Regulatory Reform 2007). This was expected to lead to the compulsory closure of around 2500 branches from a national coverage of approximately 14 000 over an 18-month period starting from the summer of 2007. The justification for this action was that the current branch network

Mitchel Langford and Gary Higgs

had simply become unsustainable. Developments in communications technology such as the widespread use of e-mail, the direct payment of state benefits, and the provision of online services for banking and other retail activities had resulted in some 4 million fewer customer visits each week to the UK post office branch network over a 2-year period. Correspondingly, network losses each week had risen from £2 million in 2005–06 to £4 million in 2006-07 (Department for Business, Enterprise and Regulatory Reform 2007). Closure proposals were developed in conjunction with organisations such as Postwatch, which had statutory responsibilities to monitor the number and location of post offices, and were outlined within area plans that subdivided the UK into over 40 regions based on groupings of parliamentary constituencies. In recognition of the importance of accessibility in maintaining social inclusion and minimising social deprivation, the closure plans were developed within a government-imposed framework of minimum geographical access criteria. These were clearly aimed at maintaining reasonable access to post offices nationally, as well as including clauses to help protect accessibility in rural and deprived urban areas. The aim of this paper is to examine the implications of the Network Change Programme using a GIS–based analysis. Branch closures inevitably reduce postal service accessibility, so our concern here is not on absolute accessibility levels per se. Instead we focus on examining whether the Network Change Programme meets its own criteria for maintaining minimum levels of acceptable service, on the extent to which the population base has been impacted by the programme including issues such as increased travel distance, loss of choice and potential environmental impacts, and finally whether there is any evidence for unequal service erosion between rural and urban communities. Wales was chosen as the focus of this study because it contains some of the most remote parts of the UK and contains higher proportions of some of the most vulnerable groups likely to be disproportionately affected by the closure programme. Recent evidence presented to the National Assembly for Wales (2008) indicated that 76 per cent of people over the age of 65 used their post office at least once a week (a figure that rose to 81% in the predominantly rural county of Gwynedd). This suggests that branch closures could have far-reaching consequences for Welsh communities where

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

there is strong support amongst consumers for the post office network (Consumer Focus Wales 2010). The rest of this paper is structured as follows. The next section provides an overview of the Network Change Programme within a policy context and reviews some of the previously published studies of UK post office network changes that have used GIS methodologies. The third section details a comprehensive GIS-based analysis of the Network Change Programme as it impacts on Wales. This work encompasses the use of a variety of spatial analytical tools including point-in-polygon analysis, network distance routing and network service catchments. The findings are discussed in the fourth section, in particular with reference to any variations in potential impact between rural and urban dwellers, before we outline further research opportunities that may arise from combining the use of spatial analytical techniques with qualitative studies. This is couched within the general framework of the use of such combined approaches to investigate spatial patterns in the provision of public (and private) services to potentially vulnerable population groups. The final section lists the main conclusions drawn from this case study.

Policy background The Network Change Programme The post office branch network in the UK has declined in numbers every year since the early 1960s, often in response to the retirement of subpostmasters ⁄ mistresses who were subsequently not replaced. In addition, between 2002 and 2005 Post Office Ltd implemented the Urban Network Reinvention Programme aimed at rationalising the branch network in urban areas, and resulting in the closure of approximately 2500 urban sub-post offices (NERA Economic Consulting 2005). Although intended to create a ‘robust network’ (House of Commons Trade and Industry Committee 2007, 9), this was essentially a straightforward reduction of urban capacity based on simple straight line distance criteria (i.e. that 95 per cent of the urban population should remain within 1 mile of their nearest branch). The Network Change Programme represents a further cumulative reduction in the branch network capacity, but is intended as a more strategic restructuring that includes both rural and urban areas and that uses new access criteria set out by government. To guide the branches chosen

587 Table I Access criteria under the Network Change Programme 1

2

3 4 5

99% of the national population should remain within 3 miles, and 90% should remain within 1 mile, of their nearest post office branch 99% of the total population in ‘deprived urban areas’ should remain within 1 mile of their nearest post office branch 95% of the total urban population should remain within 1 mile of their nearest post office branch 95% of the total rural population should remain within 3 miles of their nearest post office branch 95% of the population resident within each postcode district boundary should remain within 6 miles of their nearest post office branch

Source: Adapted from Department for Business, Enterprise and Regulatory Reform 2007, 16)

for closure, five national minimum access criteria were specified (Table I). These aimed to ensure, for example, that 99 per cent of the population would remain within 3 miles of their nearest branch outlet and 90 per cent within 1 mile. In urban areas 95 per cent of the population were to remain within 1 mile of their nearest branch, while 95 per cent of the rural population should remain within 3 miles. These targets included a possible downgrading of full branch provision to ‘outreach’ services in communities (these include mobile and home services or those partnered with other types of activities such as pubs and village halls). Geographical Information Systems were used by Post Office Ltd in the development of area plans that outlined the branches to be closed or replaced by an outreach service. This was supplemented by further GIS analysis carried out independently by Postwatch (National Audit Office 2009, 19) and resulted in some cases in area plans being amended prior to a local consultation process taking place. The social and economic impacts of post office closures have formed the basis of several recent enquiries and reports (see for example, NERA Economic Consulting 2006; House of Commons Business and Enterprise Committee 2008a 2008b). The first of the House of Commons Business and Enterprise Committee reports (2008a, 13) confirmed that the geographic access criteria are national and so may not apply at the local level, and furthermore that ‘we do not know the extent to which the national criteria are met in each local area plan’. The report further suggests that the Committee expect that ‘Post Office Ltd, as far as possible, take the geographic access criteria into account at local

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

588

area plan level, not just meet them nationally’. The government response to this was that access criteria were devised to maintain a national network of post offices but that there was to be an underpinning minimum coverage at the local level based on postcode districts. They draw on information provided in the local area plans to suggest that on average for 90 per cent of customers the proposals would mean no change to the post office they use, and that 99 per cent would see no change or would remain within 1 mile by road of an alternative branch (House of Commons Business and Enterprise Committee 2008b, 23). Post Office Ltd confirmed in the same report that it does not measure national criteria at the local level but that ‘it does monitor the impact of each local area plan proposal to ensure that overall, the network remains compliant with the National Access Criteria laid down by Government’ (House of Commons Business and Enterprise Committee 2008b, 31). It also highlighted the local criterion that 90 per cent of the population in each Postcode District must be within 6 miles of their nearest branch. It was also argued that whereas compliance with the nationally set criteria was based on ‘as the crow flies distance’, decisions at local level took into account network distances between existing branches to ‘reflect the real journeys of its customers’. Local issues were claimed to be accounted for, including the availability and timing of public transport, physical accessibility of branches and demographic characteristics of the local population. Other factors said to have influenced the programme include whether a post office is the last shop in the village and the degree of deprivation of the area in which a branch is located. Nevertheless there has been a great deal of controversy over the closure of numerous post offices, culminating in local community campaigns and petitions and a high media profile regarding the social impacts of the programme (Consumer Focus Wales 2010).

The use of GIS to analyse changes in post office provision Previous studies have demonstrated how spatial analytical approaches can be used to examine changes in post office provision (Higgs and White 1997; White et al. 1997). These have identified the socio-economic characteristics of communities likely to be impacted by changes in postal service provision following previous rounds of re-configuration since the 1970s. It was demonstrated that such

Mitchel Langford and Gary Higgs

changes could be monitored using GIS techniques and that accessibility measures could potentially be incorporated into government-based indicators of disadvantage in order to identify communities that had lost key services such as post offices. GIS analysis has subsequently been used in a number of reports to highlight gaps and changes in provision; the Commission for Rural Communities State of the countryside report for England (2008), for example, found that the percentage of rural residents that were further than 2 kilometres from a post office rose from 14.4 per cent in 2000 to 16.7 per cent in 2008, but with differences in spatial patterns evident across England. Similarly, drive-times to post offices were used to report that in Wales, prior to the closure programme in January 2005, just over 95 per cent of households were within 5 minutes and that only 0.5 per cent of households were more than 10 minutes from their nearest branch (Welsh Assembly Government 2006). Since the mid 1990s many local authorities have invested significantly in GIS technologies and several, including Cambridgeshire and Powys County Councils, have used these to monitor impacts and make counter-arguments to proposed closures in their local areas. These authorities were able to use GIS to incorporate localised data sources relating to deprivation patterns, spatial patterns of households that regularly use post office services (and who would therefore be more adversely affected by closure), local plans related to areas of future housing and predicted population growth, public transport network timetables and even maps of those parts of the road network liable to periodic flooding. Such studies were used to provide alternatives to Post Office Ltd’s model of nearest alternative post office following its proposed changes (Beetlestone et al. 2008; Soper 2009). In the case of Cambridgeshire, this led to a significant reduction in the number of branches identified for closure and justified the use of GIS software and up-to-date localised knowledge and data. In contrast, Comber et al. (2009) used GIS to examine whether national access criteria specified in the Network Change Programme were met for a local authority in the East Midlands. Their network analysis of distances between post office locations (before and after closure) and Lower Super Output Area (LSOA) population-weighted centroids found that only one criterion was met in a particular English county. They went on to use an optimisation model to design alternative spatial configurations

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

589

of branches that improved access for the population as a whole and for those population groups most vulnerable to closures or most likely to use postal services. This work reinforces and extends that of Comber et al. (2009) in a number of ways. Firstly, we conduct a much more extensive study in terms of the geographical area covered, the corresponding size of the affected population, and the number of post office branches and closures considered. Secondly, this analysis is conducted at a finer spatial scale by modelling population at the Output Area (OA) level. Thirdly, our criteria for establishing ‘urban deprived’ populations exactly matches that used by Post Office Ltd (i.e. the most deprived 30% of LSOAs in Wales). In contrast, Comber et al. used the 95th percentile of deprivation score, which does not match the 85th percentile adopted by Post Office Ltd within England. Finally, in addition to evaluating government national access criteria we test further accessibility statements made within individual area plan documents, and go on to consider other aspects of accessibility beyond nearest branch analysis such as choice reduction, travel distances and environmental impacts. A

A GIS-based analysis of the Network Change Programme in Wales Study area and datasets Our analyses are conducted across all of Wales, covering an area of approximately 80 310 square miles. The UK Census 2001 reports a population of approximately 2.9 million and its spatial distribution is modelled using population-weighted centroids supplied by the Office for National Statistics (ONS). A single representative point is provided for each OA, which is the finest level of census dissemination unit in the UK. In 2001 there were 9769 OAs across Wales, with an average population count of 297. In order to conduct some of the analyses that follow, it was necessary to classify OAs as being either ‘urban’ or ‘rural’ in character. Again we utilised an official ONS dataset, whereby this is determined on the basis of whether the majority of the population falls inside a settlement with a population size of 10 000 or more (see Bibby and Shepherd 2004; DEFRA 2005). Across Wales, 6242 OAs are defined as urban and 3527 as rural. Figure 1 (panel A) maps the population-weighted centroids C

B

D

Figure 1 Panel A: the distribution of Output Area population weighted centroids in Wales. Panel B: the distribution of post office branches before and after the Network Change Programme. Panel C: the five area plan regions within Wales. Panel D: an expanded view of the area depicted in Panel C of Output Area centroids, post office branches and Meridian road network Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Mitchel Langford and Gary Higgs

590

and the expanded view in panel D provides some indication of the spatial precision afforded by this dataset. Although exact details of the methodology and datasets employed by Post Office Ltd in its analyses of accessibility were not available to us, there is sufficient evidence in the published documentation to lead us to believe that the same datasets were utilised. Using documentation published online (Post Office Ltd 2009), we constructed in ArcGIS a comprehensive spatial database depicting the location of every post office branch in Wales prior to the Network Change Programme (Figure 1, panel B). Each branch carried an attribute to record its status with respect to the closures programme – either ‘retained’ or ‘closed’. Retained branches were further subdivided into those that would continue to offer a traditional full-time service, and those targeted to become an outreach service. Outreach services may take on any one of four formats: mobile van, hosted, partner and home service. All provide access to core postal services but with restricted opening hours (except partner services, which operate whenever the allied retail outlet is open, and therefore can offer longer opening hours than a traditional post office). Although opening times of outreach branches are published, these vary enormously; some are open for just a single hour per week, others for several hours on several days, others still for 8 hours most days of the week. Given the range of possibilities, it was felt too complex to build this detail into our model; instead, we simply record that part-time or non-traditional access hours are offered. Our GIS analyses required the construction of a network dataset to model the road transport system using topologically connected edges and intersection nodes. The OS MeridianTM 2 dataset (Edina 2009), acquired under an academic licence from Edina ⁄ Digimap, was used to create this. OS MeridianTM 2 is a vector database of approximately 1:50 000 scale Table II

Population Retained (Full time) (Outreach) Closed Total

from which motorway, A-road, B-road and minor road features were extracted and clipped to the national boundary of Wales. We did not have access to junction details so roads were assumed to have full connectivity at any intersection node. Therefore, the analyses that follow could over-estimate accessibility, since in reality turn restrictions and situations where roads are spatially coincident but non-connecting (i.e. flyovers and underpasses) will reduce true network connectivity. Inaccuracies could also arise from the fact that, although post office branches are located on or very close to the road network, the same is not necessarily true for OA population-weighted centroids. ArcGIS automatically projects these points onto the nearest orthogonal position on the road network to enable network analysis to proceed.

Preliminary analysis: overview of changes Table II shows summary statistics for post office branches located within each of the five area plan regions covering Wales (as depicted in Figure 1, panel C). It can be seen that 157 branches closed as a result of the Network Change Programme, with more than 19 per cent of the pre-2008 branch network either closed or replaced by an outreach service. However, there are noticeable regional variations with the urbanised southern regions relatively worse off in terms of absolute number of closures, although when the population base being served is taken into account North Wales also looks to be badly affected. The impacts of the Network Change Programme can be summarised using a simple polygon-containment analysis. Table III shows the number of branches closed or replaced by an outreach service, differentiated by the urban ⁄ rural status of the OA within which they reside. This simple GIS analysis suggests rural areas have suffered the highest losses, particularly in terms of access to ‘traditional’ full-time services. These figures equate to a loss of 0.4 branches per

Summary statistics of the Network Change Programme by area plan in Wales Newport and Gwent Valleys

Cardiff and Glamorgan Valleys

Central Wales

South and West Wales

North Wales

WALES

536 181 127 (124) (3) 24 (16%) 151

711 093 154 (154) (0) 30 (16%) 184

200 092 154 (130) (24) 12 (7%) 166

789 656 292 (266) (26) 39 (12%) 331

664 374 271 (258) (13) 52 (16%) 323

2 903 085 998

157 (14%) 1155

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

591

Table III Branch closures by urban ⁄ rural status of enclosing Output Area

Rural Urban

Closed

Outreach

82 75

66 0

10 000 urban residents, but 0.79 branches per 10 000 rural residents. Furthermore, if outreach branches are considered to represent a ‘loss of service’, then the latter figure rises to 1.43 branches per 10 000 rural residents. So, on the basis of a simple polygon-containment accessibility metric, rural areas have been impacted more than urban areas.

Evaluation of national access criteria To explore more fully the spatial patterns and impacts of branch closures, sophisticated network analysis of accessibility to post offices was undertaken using ArcGISTM Network Analyst. Our first analysis aimed to evaluate whether the national minimum access criteria prescribed by the UK Government (listed in Table I) are met within the confines of Wales. Using the Closest Facility tool, road distance to the nearest post office was computed for every OA before and after the Network Change Programme. The operation is depicted schematically in Figure 2. In the latter case, we further differentiate between outreach facilities being included and being excluded from the analysis. By excluding them we effectively evaluate accessibility to a ‘fulltime’ service only. To evaluate criteria (2), (3) and (4), we used the ONS urban ⁄ rural classification discussed earlier. To evaluate criterion (2), we must further identify ‘deprived’ urban OAs. This was achieved by selecting those OAs that fell within the most deprived 30 per cent of LSOAs in Wales according to the Welsh Index of Multiple Deprivation Score 2005 (Welsh Assembly Government 2005). To evaluate criterion (5), we conducted pointin-polygon analysis of population-weighted centroids within postcode district boundaries. Postcode boundaries are subject to change over time and no details are available regarding which specific data were used by Post Office Ltd. We employed 2001 CAS postcode sector boundaries available from UK Borders ⁄ Edina, subsequently dissolved in ArcGIS to create district-level boundaries. The results are summarised in Table IV. With respect to criterion (1), the accessibility requirement

Figure 2 Schematic showing the computation of network distances. The population resident at Output Area (A) must adopt a different route through the road network due to its nearest post office branch closing as a result of the Network Change Programme. In contrast, residents at Output Area (B) are unaffected by the Programme as their nearest post office branch remains open

is not met before any branch closures are considered, and so can only deteriorate further as a result of the Network Change Programme. The same situation is found with respect to criterion (2), with significantly less than 99 per cent of the total urban deprived population having access to a branch within 1 mile network distance. In absolute terms,

Table IV Assessment of national access criteria Before closure (%)

After closure (including outreach) (%)

After closure (excluding outreach) (%)

miles mile

98.39 87.58

98.32 85.30

96.96 84.56

mile

97.81

96.49

96.49

mile

94.69

92.69

92.69

miles

95.49 Satisfied in all postcode districts

95.30 Satisfied in all postcode districts

91.48 Failed in 9 postcode districts

Criterion (1) £3 £1 (2) £1 (3) £1 (4) £3 (5) £6

miles

Notes: Bold figures indicate failure to meet national target figures. Total urban deprived population=697 160; total urban population=1 864 269; total rural population=1 036 815

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Mitchel Langford and Gary Higgs

592

this equates to approximately 2600 individuals before the closures, and 3500 individuals after, that have inadequate accessibility according to this stipulation. A similar pattern is repeated for criterion (3) relating to branch accessibility for the general urban population. Our analysis estimated that an additional 37 416 individuals were left with ‘inadequate’ access to a post office branch over and above the 5600 or so that did not meet this criterion prior to the closures. In contrast with the earlier finding based on a simple polygon-containment accessibility metric, our examination of criterion (4) indicates that the rural population is somewhat better served when judged using the government-prescribed accessibility criterion. More than 95 per cent of the rural population is able to access a branch within 3 miles both before and after the Network Change Programme, provided that a part-time service is deemed acceptable. However, the importance of the outreach branches in maintaining acceptable accessibility levels is made very clear. If these are excluded from the analysis the situation deteriorates considerably, leaving more than 88 000 rural residents living further than 3 miles from a full-time service delivery point. This is 36 500 more than is ‘allowable’ under the 95 per cent rule. Criterion (5) measures accessibility at a more local scale (postcode district), while adopting a much less stringent accessibility requirement. Before the closures only two postcode districts had any population located beyond the 6 mile limit, equating to 2.0 per cent and 1.8 per cent of their respective population totals – both well within government stipulations. The situation remains unchanged after the closures, provided outreach branches are taken into consideration. Once again the importance of outreach in allowing government accessibility targets to be met in rural areas is quite clear. When excluded, we calculate 14 postcode districts to have a proportion of their population residing more than 6 miles from a full-time service. Of these, nine would fail the 95 per cent rule and in one case spectacularly so; if the mobile outreach service in postcode district SY19 (operating 9 hours per week) is discounted, only 21 per cent of its population lies within 6 miles of a post office branch. Overall, it is clear from these analyses that national minimum accessibility criteria for post offices are not being met within the confines of Wales.

Accessibility within area plan regions Post Office Ltd has produced reports for each of the 42 area plan programme delivery regions, documenting the numbers of post offices to be closed, the specific branches affected and the arrangements made for outreach services to operate. Within these documents claims are made with regard to the accessibility impacts of the Network Change Programme in each specific area (these are additional to the national criteria reported and analysed above). Once again, the exact methodology used to provide the evidence to support these statements is unclear, but it is reasonable to assume it is essentially the same as that used for the national criteria. These additional statements provide generally reassuring evaluations of the minimal impact that Network Change will have in each area plan region. For example, the Cardiff and Glamorgan Valleys area plan document states that More than 99.7% of the population will see no change to the branch that they currently use or would remain within one mile of an alternative outlet. (Post Office Ltd 2008, 5) (1)

and Of the 9.1% of customers who use a branch that will close, 8.8% would have access to Post Office services within one mile as measured by road distance (2.7% being within half a mile of their nearest alternative). (Post Office Ltd 2008, 19) (2)

Comparable claims are made for each of the five area plan regions in Wales, as summarised in Table V. This table also displays our results using ArcGIS Network AnalystTM to test the validity of these statements. In particular, we examined the validity of statements such as (2) listed above. Some care is needed in determining the precise meaning of these statements, since some ambiguity in the interpretation of the ‘within one mile’ and ‘within half a mile’ clauses is possible. Specifically, they do not make clear from what starting point these distances are measured. Our initial interpretation (which we suspect would be the same as most readers) was that they imply the majority of residents who lose their current nearest branch would travel no more than 1 mile (or half a mile) from their home to access an alternative post office. Thus we first computed the percentage of population that would see no change to the branch they currently use or would have access to an alternative branch within 1 mile of their OA centroid. The respective figures

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

593

Table V Assessment of accessibility statements made in area plan documents for Wales

apc1

apc2

apc3

apc4

(claimed) Inc. outreach Exc. outreach (claimed) Inc. outreach Exc. outreach (claimed) Inc. outreach Exc. outreach (claimed) Inc. outreach Exc. outreach

Newport and Gwent Valleys

Cardiff and Glamorgan Valleys

Central Wales

South and West Wales

North Wales

(7.0%) 9.45% 10.54% (6.7%) 8.85% 9.15% (2.0%) 5.77% 5.93% (99.6%) 99.4% 98.6%

(9.1%) 10.08%  (8.8%) 9.54%  (2.7%) 7.26%  (99.7%) 99.6% 

(10.7%) 6.43% 15.88% (9.5%) 4.64% 5.59% (6.2%) 3.07% 3.65% (98.8%) 97.5% 98.6%

(6.1%) 7.60% 10.12% (5.5%) 6.87% 7.33% (2.4%) 4.72% 4.84% (99.4%) 99.3% 97.2%

(8.5%) 12.45% 13.58% (7.3%) 10.77% 11.89% (3.7%) 7.80% 8.93% (98.8%) 98.3% 89.7%

Notes: apc1 = % of population whose nearest Post Office has closed; apc2 = % of population within 1 mile of the nearest alternative Post Office; apc3 = % of population within ½ mile of the nearest alternative Post Office; apc4 = % of population that either experience no change in service or are within 1 mile of the nearest alternative Post Office

were: 97 per cent (Newport & Gwent), 98.2 per cent (Cardiff and Glamorgan), 95.6 per cent (Central & East), 97.4 per cent (South & West) and 95 per cent (North) – all of which fall considerably short of the claimed figures listed under apc4 in Table V. In order to better match Post Office Ltd figures, we then considered the possibility that these statements should be interpreted in terms of how much further a given population must travel if their nearest post office closes. In other words, we computed the percentage of population that would have to travel no more than 1 mile (or half a mile) further to reach a post office than the journey length undertaken before the closure of their nearest branch. It is these calculations that are reported in Table V, but even with this more generous interpretation we still could not match any of the published claims in any of the five area plan regions of Wales. Furthermore, in all regions, except Central Wales, we estimated a higher percentage of resi-

dent population would suffer the closure of its nearest post office branch than the published figure (i.e. apc1 in Table V). Moreover, if the argument is accepted that those whose nearest branch was replaced by an outreach service are not unaffected by the programme, then the percentage of population impacted ranges from just over 10 per cent in South & West Wales to almost 16 per cent in Central Wales. Further insight into the implications and impacts of closures in each region are given by the analysis presented in Table VI. For those communities affected by closure, the average distance that must be travelled to access a branch before and after the programme has effectively doubled in all area plan regions in Wales. The average distance to a replacement service delivery point is also notably higher in all cases than the average distance travelled by those communities unaffected by the changes. The maximum additional travel distance required to reach a branch was computed as 2.47 miles for an

Table VI Impacts of the Network Change Programme on travel distances (in miles) Newport and Gwent Valleys

Distance Distance Distance Distance

before after increase for unaffected

Cardiff and Glamorgan Valleys

Central Wales

South and West Wales

North Wales

Ave

Max

Ave

Max

Ave

Max

Ave

Max

Ave

Max

0.47 0.94 0.47 0.57

3.28 3.46 1.99 5.61

0.36 0.76 0.39 0.47

2.32 3.32 1.91 3.71

0.75 1.51 0.76 0.95

3.70 4.18 2.32 6.16

0.51 1.00 0.48 0.58

3.35 3.51 2.22 5.36

0.53 1.05 0.51 0.60

3.86 4.18 2.47 5.80

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

594

Mitchel Langford and Gary Higgs

Figure 3 Schematic showing the computation of choice within Output Area service areas. Residents in urban Output Area (A) experience the loss of one post office branch within 1 mile network distance of the Output Area’s centroid, but retain a choice of two other branches. Residents in rural Output Area (B) lose access to their nearest branch, but retain one post office within 3 miles network distance of the Output Area’s centroid. Residents in rural Output Area (C) lose access to any post office branch within 3 miles network distance of the Output Area’s centroid. Note that service area catchments shown here are highly stylised – in reality they possess complex shapes determined by the local road network pattern

OA in North Wales, but in all area plan regions the maximum additional travel was close to, or above, 2 miles. However, it should be noted that there were communities unaffected by the programme that must travel considerably further than this to reach a branch (for example, over 6 miles for the worst case in Central Wales). All these figures are based on an analysis that includes outreach services. To gain access to a full-time branch, the average distance increases for those affected by the programme are 0.71 miles (Newport & Gwent), 1.89 miles (Central & East), 1.01 miles (South & West) and 0.72 miles (North). Furthermore, the maximum travel distance required to reach a branch then exceeds that for the unaffected communities in all cases (except Cardiff and Glamorgan where, with no Outreach branches, there is no change).

Accessibility, choice and environmental impacts within communities As stated previously, accessibility is a complex phenomenon and distance to nearest facility is only one of a number of possible metrics. In order to provide a more holistic view of the impacts on availability and choice for individual communities at the OA level, as well as possible environmental consequences, we used ArcGIS Network Analyst’s

Service Area tool to construct network-based catchments. Using the OA centroids, service areas were computed based on the minimum access guidelines previously described (i.e. 1 mile for urban OAs, 3 miles for rural OAs). Then, using point-inpolygon analysis, a database was constructed to identify all branches that lay within the specified proximity limits of each OA. This allowed a detailed profile of post office facilities in the vicinity of each OA both before and after the Network Change programme to be established. The analytical procedure is depicted schematically in Figure 3. Our contention here is that people do not always behave totally rationally and therefore do not travel only to their nearest post office on every occasion. In reality there are likely to be many factors that may make them choose another nearby branch, or cause them to vary their choice from time to time. Potential reasons include: the provision or lack of parking facilities at a chosen branch; physical access to a branch; whether the branch is en-route to a workplace or to family and friends; and the proximity of a branch to other retail services. Table VII reports the number of residents with access to the specified number of full-time (i.e. outreach-excluded) branches before (by row) and after (by column) the Network Change Programme. It

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision Table VII

595

Full-time branch availability for all Output Area residents by service area No. of branches accessible after programme

No. of branches accessible before programme

0 1 2 3 4 5+ Totals

0

1

2

154 289 73 083 8464 369 0 0 236 205

0 754 208 169 021 26 668 865 0 950 763

0 0 571 110 210 779 30 440 3823 816 154

shows, for example, that over 154 000 Welsh residents had no full-time branch within government accessibility criteria before Network Change, rising to over 236 000 residents after the closures. Of the 827 000 + residents that had access to one full-time post office branch before the closures, some 73 000 will have lost this service as a result of Network Change. Furthermore, almost 8500 residents have lost access to two previously accessible full-time branches, and one particularly unlucky OA with 369 residents experienced the loss of three full-time branches within the designated proximity guidelines. Another aspect of the Network Change Programme revealed in Table VII is the reduction in choice likely to be suffered as a result of branch closures. Of the 748 500 + people who previously Table VIII

3

250 111 26 388

0 0 0 922 653 152 730

4

5+

Totals

0 0 0 0 127 508 95 997 223 509

0 0 0 0 0 287 734 287 734

154 827 748 488 270 413 2 903

289 291 595 738 466 706 085

had a choice of two full-time branches within prescribed proximity criteria, some 169 000 + have been reduced to a solitary branch. Similarly, more than 26 500 residents have seen choice diminish from three branches to one, while 865 have seen a reduction in choice from four branches to one. In aggregate, almost 197 000 residents have lost any choice in full-time post office branch within the government access criteria limits. Although less critical in nature, similar choice reduction from 3 to 2 branches (210 779 residents), 4 to 2 branches (30 440 residents) or even 5 + to 2 branches (3823 residents) are revealed in this table. Finally, a tally of the diagonal axis of this table determines the proportion of population unaffected (or alternatively, impacted in some way) by the Network Change Programme. This calculation suggests 1.95

Full-time branch availability for urban ⁄ rural Output Area residents by service area No. of branches accessible after programme 0

Rural: No. of branches accessible before programme

49 097 38 540 2756 0 0 0 90 393

0 1 2 3 4 5+ Totals

1

2

0 171 063 35 713 8592 865 0 216 234

0 0 129 974 77 752 16 322 2471 226 521

3

95 52 15 163

0 0 0 546 508 401 458

4

5+

0 0 0 0 49 358 59 838 109 200

0 0 0 0 0 231 300 231 300

Totals 49 209 168 181 119 309 1 037

097 603 443 890 053 010 096

No. of branches accessible after programme Urban: No. of branches accessible before programme

0 1 2 3 4 5+ Totals

0

1

2

105 192 34 543 5708 369 0 0 145 812

0 583 145 133 308 18 076 0 0 734 530

0 0 441 136 133 027 14 118 1352 589 635

3

155 59 10 225

0 0 0 376 145 751 275

4

5+

Totals

0 0 0 0 78 150 36 159 114 313

0 0 0 0 0 56 434 56 434

105 617 580 306 151 104 1 865

192 688 152 848 413 696 989

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

596

million residents were unaffected, but 947 100 (or 32.6% of the total population of Wales) have experienced some impact in terms of accessibility to, or choice of, full-time post office branches within government-specified distance criteria. This paints a rather different picture to the claim that 99 per cent of the population is unaffected or remains within 1 mile by road of an alternative branch (House of Commons Business and Enterprise Committee 2008b, 23). Returning to the issue of any rural ⁄ urban disparity, repeating the analysis independently for urban and rural classified OAs helps to shed further light on this question. Table VIII reveals that 49 000 rural residents and 105 000 urban residents had no access to a full-time branch inside the governmentdesignated proximity criteria before Network Change. These figures rise to approximately 90 500 and 146 000 respectively (an increase of 84% and 39%) after the closures. The proportion of population losing previous access to a full-time branch is 3.98 per cent in rural OAs and 2.17 per cent in urban OAs. Conversely, the proportion of population facing a loss of choice is relatively higher in urban than in rural OAs. Of those who had access to two branches previously, some 23 per cent are reduced to potentially accessing only one branch in urban areas, while in rural areas the corresponding figure is 21.2 per cent. The figures for a reduction from three to one branch are 5.9 per cent in urban areas and 4.7 per cent in rural areas. It would appear then that urban areas have suffered a greater reduction in choice, but that rural areas have been harder hit in terms of losing access altogether within designated proximity guidelines. By further manipulation of the database outside of the GIS using bespoke Visual Basic .NET programs, we compute that 330 000 people are left to travel to a full-time post office branch that was further away after the Network Change Programme was completed (with 58% of this affected population residing in urban areas and 42% in rural areas). The Wales-wide average increase in travel distance for those affected by the programme was 1.6 miles for the round trip, while the largest increase (Merthyr Cynog) was 8 miles. In total, 76 OAs, all designated as rural and having a combined count of 22 986 residents, have an increased journey length (on each leg of the journey) of more than 3 miles. Furthermore, 137 OAs with 41 214 residents have an increased journey length of more than 1.8 miles; and still only three of these OAs are

Mitchel Langford and Gary Higgs

classified as urban. The average increase in travel distance amongst the affected urban population was 0.39 miles per leg of the journey, while for those residing in rural areas it was 1.5 miles per leg. With respect to increased travel distances, rural residents are clearly disadvantaged compared with urban dwellers, although this is probably to be expected. By multiplying extra travel distances by the number of residents affected, we also calculate the extra ‘person-miles’ involved if each resident accessed their nearest full-time post office branch after Network Change. This figure was determined to be just over half a million miles (565 537), and represented an increase of 16.4 per cent on the preclosures value. Clearly this must cause environmental impacts in terms of both congestion and carbon emissions. For example, if we assume each person travels independently by a car with average CO2 emissions (say 170g ⁄ km), this equates to an increase of 154.7 tonnes per trip, or 8045 tonnes per year assuming a trip is made once a week. While there are clearly many assumptions in this analysis relating to, for example, the differential use of the Post Office, frequency of service use, mode of travel or primary purpose of the trip, we posit that this is indicative of the potential environmental impact attributable to the Network Change Programme and should have provided a more important component in the decision to close or change the status of post offices in the Programme.

Impacts on post office catchments Our final investigation was to examine the service areas around each post office, identifying all OAs for which any given branch is the nearest and calculating the total population served. This allowed the potential extra load on those branches that remained open after Network Change to be established. An intended outcome of the Programme was to enlarge the customer base of remaining branches to improve their future financial viability. Our analysis revealed the largest single increase to be an additional customer base of 7219 persons, while 33 branches witnessed an increase of more than 2500 persons, and 112 branches saw an increase of more than 1000 persons. However, 626 of the 933 remaining branches would see no change in customer base. Although we can identify which branches have been most affected, only Post Office Ltd can evaluate whether these are the ones most in need of an enlarged customer base to

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

remain viable. This analysis does raise important questions as to whether the potential increased loads have been fully anticipated and are manageable for post offices that remain. In terms of percentage increase, we predict for example that one branch could potentially increase its customer base by a staggering 2345 per cent under the assumptions of our analysis, and another by over 1000 per cent. A further 58 branches more than double their population base as a result of the closures. The two branches experiencing the largest percentage gains are both located in urban OAs (central Pontypridd and the outskirts of Bridgend), but amongst the 60 branches experiencing a more than doubling of their customer base, the split is biased in favour of rural branches at 36 ⁄ 24. The same bias occurs amongst the 298 branches that showed any increase in their customer base with a 170 ⁄ 128 split. This demonstrates the potential of Network Change to increase the future viability of remaining branches, particularly in rural areas.

Discussion Confirming recent preconceptions of the impacts of the Network Change Programme, preliminary analysis using simple polygon-containment accessibility metrics suggests that rural residents in Wales were hardest hit by post office closures (with 0.79 branches lost per 10 000 rural residents compared with 0.4 for urban residents). However, adopting more sophisticated network-distance based analyses paints a much more complex and often contrasting picture. It suggests that urban rather than rural residents may have been most affected by reduced accessibility due to the recent closures programme. While the proportion of population losing access to a full-time branch is higher in rural OAs (3.98% versus 2.17%), this is ameliorated to a large extent by the provision of part-time outreach services. Meanwhile, loss of choice is clearly greater in urban areas; the figures for a reduction from two accessible branches to only one are (urban) 23 per cent versus (rural) 22 per cent, and for a reduction from three branches to one are 5.9 per cent versus 4.7 per cent. Government criteria that underpin the Network Change Programme and that aim to ensure minimum accessibility levels are not met in either urban or deprived urban areas (as well as across Wales as a whole), but they are satisfied in rural OAs. A similar bias of impact on the urban population was reported by Comber et al.

597

(2008) for the county of Leicestershire. This reversal of outcome may be attributable to the greater sophistication of the network analysis tools, but perhaps also reflects the more lenient accessibility criteria imposed in rural regions (i.e. proximity within 3 miles rather than 1 mile). It should be noted that Post Office Ltd state that these accessibility criteria need only be met nationally (i.e. UK-wide) rather than for any individual country of the UK: The emphasis of our policy is to maintain a national network with national coverage. We reject therefore the proposal that the criteria above should be applied at the level of individual countries or smaller local areas. (Department for Business, Enterprise and Regulatory Reform 2007, 16)

But this assertion appears to be completely incompatible with other statements made with respect to ensuring equitability in the distribution of impacted populations: In rationalising the network, we would require Post Office Ltd to ensure that no one part of the network or no particular group of people is significantly more adversely affected than any other. We expect that rationalisation will principally affect a combination of least used branches and non-commercial branches in areas of over-provision where people can find alternative branches nearby (Department of Trade and Industry 2006, 23) Many respondents also called for parity of treatment between rural and urban areas, between countries within the UK and for parity to be maintained [...] We agree that no particular part of the network and no particular group of people should be significantly more adversely affected by closures or other changes in service provision than any other. We therefore expect that Post Office Ltd will be making roughly similar numbers of closures in rural and urban areas. We also expect that when developing detailed area plans Post Office Ltd will reflect the principle that no country within the UK and no group of inhabitants at the area plan level should be significantly more adversely affected than any other. (Department for Business, Enterprise and Regulatory Reform 2007, 18)

Post Office Ltd also reports that the closures programme will in most cases leave areas exceeding the minimum access criteria: The access criteria replace the requirement placed on Post Office Ltd to prevent avoidable closures of rural offices. But it is important to note that these criteria represent the minimum levels of accessibility and in many, if not most, cases actual coverage will be greater.

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

598 (Department for Business, Enterprise and Regulatory Reform 2007, 16)

If these national criteria are indeed met or exceeded nationally (which could only be established by further independent testing) then given the results above it follows that there must be other regions in the UK where the criteria are sufficiently exceeded so as to balance out the under-attainment within Wales. Once again the statement that no particular group of people or country should be adversely affected compared with any other appears, at best, to be contestable. We also examined claims made with respect to impacts within individual area plan regions. Notwithstanding the somewhat opaque wording of these statements, our analysis failed to match published figures even when using the most generous interpretation. Our results show that amongst those communities affected by closure, the average distance to a post office has effectively doubled in all regions in Wales. Furthermore, we argue that residents are affected by the Network Change Programme not only if their nearest branch closes, but also through broader accessibility measures such as loss of choice within a neighbourhood, or by the downgrading of a service to a part-time basis. In Central Wales, for example, almost 16 per cent of the population have lost their nearest branch or seen it replaced by an outreach service. Across Wales as a whole, around 197 000 residents no longer have a choice of branch within government-specified proximity limits. We compute that almost a third of the total population of Wales has experienced some impact in terms of accessibility to, or choice of, full-time post office branch as a result of the Network Change Programme. To gauge the full impact of the closure programme, ideally researchers require a comprehensive analysis by type of closure (for example, permanent loss of a service, change in hours of the new service and change to the type of service such as outreach). This suggests, for example, that the hours of service provided by replacement outreach services as well as their physical location should be investigated. In this study we have only considered the numbers and type of provision (full-time ⁄ outreach) of post offices, and have not investigated changes to the quality of provision, the physical access arrangements or the range of services provided. Outreach services, for example, may not provide the full range of post office services,

Mitchel Langford and Gary Higgs

although, in some cases, more services are offered than in the fixed post office branch it replaced (Consumer Focus 2009). The potential implication of such changes on different socio-economic groups should form a part of future studies in this area. Some work has already been done on the implications of potential changes in services offered by post offices in relation to wider notions of financial exclusion (e.g. Midgley 2005) and the wider social and economic importance of post offices is recognised by the UK Government through the £150 million per year Social Network Payment. Post Office policy is not devolved but in Wales grants have been made available under the Welsh Assembly Government’s Post Office Diversification Fund to develop new services within sub-post offices. However, more research is needed to explore socio-economic and demographic trends in those areas impacted by this programme. A recent report published by Consumer Focus Wales (2010), based on focus groups and in-depth interviews, highlighted the impact on vulnerable groups such as the elderly and the disabled, those reliant on public transport and people on low incomes within five communities in Wales that had experienced closures. Those who had to travel further to access their nearest post office complained of poorer experiences of using alternative branches, of a feeling of social isolation following a closure and difficulties in physically accessing postal services. We believe the analysis included in this paper contributes evidence to a wider discourse on such issues related to social exclusion and to the disproportionate impacts of service reconfigurations on specific vulnerable groups. Some research has already begun to explore the types of areas likely to be most affected by post office closures. Macintyre et al. (2008), in a survey of 42 types of facilities in Glasgow in 2005–06, found that post offices were more common in more deprived neighbourhoods and that the number of branches per 1000 people actually increased with increasing deprivation. A National Consumer Council report investigated the impacts of closures undertaken as part of the Urban Reinvention Programme (2002–06), particularly in disadvantaged urban estates in cities such as Cardiff, Nottingham and Leeds (National Consumer Council 2007). Their report identified younger groups and those on benefits as likely to be the most impacted, as

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision

well as elderly people with no access to private transport. Using a geodemographic classification at postcode sector level and changes in post office provision per household, the authors argue that such systems can be used to gauge social need and should feed into any future closures programme. Public service provision needs to be seen in the context of wider social justice and social inclusion agendas and in particular the goal of giving people greater access to essential services. Recent policy reports continue to focus on problems of accessibility and on the wider implications of changes in the provision of essential services such as post offices, food shops and health facilities. A recent Commission for Rural Communities consultation exercise, for example, drew attention to the importance of access to services ⁄ facilities as a major element of disadvantage particularly in relation to population trends in rural areas (Commission for Rural Communities 2005). Leyshon et al. (2008) used GIS to analyse bank and building society closures in Britain between 1995 and 2003 in relation to on-going debates regarding processes of financial exclusion. Based on a comparison of the spatial distribution of closures with an index of multiple deprivation and a geodemographic classification, they found that branches tended to be closed at a faster rate in more deprived areas. This, in turn, suggests an ‘uneven geography’ in financial services provision in Britain (although other studies quoted by Leyshon et al. did not find such a trend). Future research should investigate whether such trends are replicated in the provision of postal services. Finally, at a time when the government is promoting policies and programmes to reduce private transport emissions and help develop a low carbon economy, a dimension that has been relatively under-researched is the environmental consequences of proposed service provision changes. In this paper we have provided provisional predictions of the likely consequences of the Network Change Programme in Wales and call for more research to model the impacts of the increased journeys needed to access postal services following branch network reconfiguration. In particular, attention should be given as to whether the environmental costs have exceeded the initial savings outlined in the closure programme. GIS could clearly be used to help estimate the additional time and effort needed to access post office branches following any closure programme under different

599

modelling scenarios and this, in turn, could form the basis of further investigation in relation to the re-configuration of other public services such as health, retail and leisure facilities.

Conclusions The Network Change Programme is expected to have resulted in 2435 post office branch closures across the UK (National Audit Office 2009). In Wales some 223 branches have either been lost altogether or replaced by part-time outreach services as a result of this latest strategic reshaping of the post office network. This study utilised a detailed spatial database and GIS analytical tools to examine some of the potential impacts of these changes. Results indicate that three of the five national criteria specified by government to ensure that minimum standards of accessibility are maintained are not met within Wales. Furthermore, we were unable to match further claims made in local area plan documents concerning the percentage of population that would either see no change in service or who would remain within 1 mile of an alternative branch. We also consider wider issues such as the reduction of choice in available branches, the potential extra load on those branches that remain open and possible environmental impacts of the programme. Our analysis suggests some 197 000 residents have lost any effective choice in full-time post office branch within government-specified proximity criteria, that more than 60 remaining branches would more than double their potential customer base, and that there could be a significant increase in CO2 emissions attributable to the extra travel now needed in order to reach a post office. We estimate that more than 32 per cent of the total population of Wales has been impacted in some way by the Network Change Programme. However, with regard to the debate concerning inequitable impact between urban and rural communities, our findings depict a rather complex picture. Branch closures per head of population are worst in rural areas, yet national rural access criteria are met, while urban access criteria are not. Loss of choice appears to be greatest in urban OAs, but increased travel distances caused by this network reconfiguration are greater for rural OAs. Of the 137 OAs whose increase in journey length is more than 1.8 miles, only three were classified as urban. Therefore our analysis contradicts the perception that it is mainly rural areas that have been mostly affected, but suggests that whether

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

600

urban or rural dwellers are deemed to be most impacted depends on the specific metric adopted. Finally, as an avenue for further research, spatial analytical approaches could be used to examine accessibility in relation to the opening times of the re-configured service in relation to, for example, public transport provision. More generally, the type of analysis conducted here has the potential to integrate with qualitative approaches to examine, for example, how the potential access measures developed in this initial investigation relate to both detailed service utilisation patterns for different social groups as well as the perceptions of users prior to, and following, service change. Our future research will incorporate detailed analysis of the characteristics of those groups most dependent on post office provision and thus those most likely to be impacted by changes in service provision. We contend that this study provides an important contextual backdrop with which to carry out in-depth community-level studies of the real impact of such a closure programme on potentially vulnerable groups and points the way to combined approaches to address issues of public service provision and accessibility.

References Apparicio P and Seguin A 2006 Measuring the accessibility of services and facilities for residents of public housing in Montreal Urban Studies 187–211 Beetlestone S, Greaves D and Morgan G. 2008 The impact of post office closures on local neighbourhoods in Powys The British Urban and Regional Information Systems Association Newsletter 177 2–8 Bibby P and Shepherd J 2004 Developing a new classification of urban and rural areas for policy purposes – the methodology (http://www.statistics.gov.uk/geography/ downloads/Methodology_Report.pdf) Accessed 7 June 2009 Comber A, Brunsdon C and Green E 2008 Using a GISbased network analysis to determine urban greenspace accessibility for different ethnic and religious groups Landscape and Urban Planning 86 103–14 Comber A, Brunsdon C, Hardy J and Radburn R 2009 Using a GIS-based network analysis and optimisation routines to evaluate service provision: a case study of the UK Post Office Applied Spatial Analysis and Policy 2 47–64 Commission for Rural Communities 2005 Taking forward our study of rural disadvantage CRC, Cheltenham Commission of Rural Communities 2008 State of the countryside 2008 CRC, Cheltenham Consumer Focus 2009 Within reach? Consumer reactions to new post office outreach services Consumer Focus, London

Mitchel Langford and Gary Higgs Consumer Focus Wales 2010 Post office closures: impact of the Network Change Programme Consumer Focus Wales, Cardiff Currie G, Richardson T, Smyth P, Vella-Brodrick D, Hine J, Lucas K, Stanley J, Morris J, Kinnear R and Stanley J (2009) Investigating links between transport disadvantage, social exclusion and well-being in Melbourne – preliminary results Transport Policy 16 97–105 Department for Business, Enterprise and Regulatory Reform 2007 The post office network: government response to public consultation BERR, London (http://www.berr. gov.uk/files/file39479.pdf) Accessed 7 June 2009 Department for Environment, Food and Rural Affairs 2004 Rural strategy 2004 DEFRA, London Department for Environment, Food and Rural Affairs 2005 Rural and urban area classification 2004: an introductory guide (http://www.defra.gov.uk/rural/rural stats/rural-defn/Rural_Urban_Introductory_Guide.pdf) Accessed 7 June 2009 Department of Health 2002 Improvement, expansion and reform: the next 3 years: priorities and planning framework, 2003–2006 Department of Health, London Department of Trade and Industry 2006 The Post Office Network: A consultation document DTI, London. (http:// www.bis.gov.uk/files/file36025.pdf) Accessed 7 June 2009 Edina 2009 Digimap collections: OS Meridian 2 (http:// edina.ac.uk/digimap/description/products/meridian. shtml) Accessed 7 June 2009 Gulliford M and Morgan M 2002 Access to health care Routledge, London Halden D 2002 Using accessibility measures to integrate land use and transport policy in Edinburgh and the Lothians Transport Policy 9 313–24 Higgs G and White S 1997 Changes in service provision in rural areas: part 1: the use of GIS in analysing accessibility to services in rural deprivation research Journal of Rural Studies 13 441–50. House of Commons Business and Enterprise Committee 2008a Post Office Closure Programme: third report of session 2007–2008 Volume 1: Report together with formal minutes HC292-I and Volume II: Oral and written evidence HC292-II The Stationary Office, London House of Commons Business and Enterprise Committee 2008b After the Network Change Programme: the future of post office network HC577 The Stationary Office, London House of Commons Trade and Industry Committee 2007 Restructuring the post office network HC593 The Stationary Office, London Jones A, Hillsdon M and Coombes E 2009 Greenspace access, use and physical activity: understanding the effects of area deprivation Preventative Medicine 49 500–5 Langford M, Higgs G, Radcliffe J and White S 2008 Urban population distribution models and service accessibility estimation Computers, Environment and Urban Systems 32 66–80

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Accessibility and public service provision Larsen K and Gilliland J 2008 Mapping the evolution of ‘food deserts’ in a Canadian city: supermarket accessibility in London, Ontario, 1961–2005 International Journal of Health Geographics 7 Lei T and Church R 2010 Mapping transit-based access: integrating GIS, routes and schedules International Journal of Geographical Information Science 24 283–304. Leyshon A, French S and Signoretta P 2008 Financial exclusion and the geography of bank and building society branch closure in Britain Transactions of the Institute of British Geographers 33 447–65 Macintyre S, Macdonald L and Ellaway A 2008 Do poorer people have poorer access to local resources and facilities? The distribution of local resources by area deprivation in Glasgow, Scotland Social Science and Medicine 67 900–14 Martin D, Jordan H and Roderick P 2008 Taking the bus: incorporating public transport timetable data into health care accessibility modelling Environment and Planning A 40 2510–25 McEntee J and Agyeman J 2010 Towards the development of a GIS method for identifying rural food deserts: geographic access in Vermont, USA Applied Geography 30 165–76 Midgley J 2005 Financial exclusion, universal banking and post offices in Britain Area 37 277–85 National Assembly for Wales 2008 Poverty and deprivation in rural Wales National Assembly for Wales, Rural Development Sub-Committee, Cardiff National Audit Office 2009 Oversight of the Post Office Network Change Programme: report by the Controller and Auditor General HC558 The Stationary Office, London National Consumer Council 2007 Post office closures 2002 to 2006: lessons for 2007 to 2009 NCC, London NERA Economic Consulting 2005 Review of the post office urban reinvention programme: a final report for PostComm NERA Economic Consulting 2006 Post office urban reinvention – one year on: a final report for Postcomm Noble M, Wright G, Smith G and Dibben C 2006 Measuring multiple deprivation at the small-area level Environment and Planning A 38 169–85

601 Omer I 2006 Evaluating accessibility using house-level data: a spatial equity perspective Computers, Environment and Urban Systems 30 254–74 Post Office Ltd 2008 Network Change Programme: Cardiff and Glamorgan Valleys area plan (additional consultations, decision booklet) Post Office, London Post Office Ltd 2009 Post Office Network Change Programme Official site (http://www.postoffice.co.uk/ portal/po/content2?catId=57600693&mediaId=57600697) Accessed 7 June 2009 Social Exclusion Unit 2003 Making the connections: transport and social exclusion Office of the Deputy Prime Minister, London Soper M 2009 SOS – save our stores: using research to help limit post office closures in Cambridgeshire Local Authorities Research and Intelligence Association Newsletter 90 18–19 Talen E 2003 Neighbourhoods as service providers: a methodology for evaluating pedestrian access Environment and Planning B 30 181–200 Wang F and Luo W 2005 Assessing spatial and nonspatial factors for healthcare access: towards an integrated approach to defining health professional shortage areas Health & Place 11 131–46 Welsh Assembly Government 2005 Welsh index of multiple deprivation 2005: summary report Welsh Assembly Government, Cardiff Welsh Assembly Government 2006 Evaluation of the Post Office Development Fund, social justice and regeneration Research report SJRRR 2 ⁄ 06 Welsh Assembly Government, Cardiff White S, Guy C and Higgs G 1997 Changes in service provision in rural areas: part 2: changes in post office provision in mid Wales: a GIS based evaluation Journal of Rural Studies 13 451–65. Witten K, Exeter D and Field A 2003 The quality of urban environments: mapping variation in access to community resources Urban Studies 40 161–77 Yang D, George R and Muller R 2006 Comparing GIS-based methods for measuring spatial accessibility to health services Journal of Medical Systems 30 23–32

Trans Inst Br Geogr NS 35 585–601 2010 ISSN 0020-2754  2010 The Authors. Transactions of the Institute of British Geographers  2010 Royal Geographical Society (with the Institute of British Geographers)

Suggest Documents