Measuring deprivation in Scotland: developing a long-term strategy: interim report

Measuring deprivation in Scotland: developing a long-term strategy

Interim report

Nick Bailey, John Flint, Robina Goodlad, Mark Shucksmith, Suzanne Fitzpatrick & Gwilym Pryce

May 2003 http://www.scrsj.ac.uk/deprivation

The Report

This report has been written as part of a project to develop a long-term strategy for measuring

deprivation in Scotland. It is an Interim Report designed to stimulate wider discussion of

these issues and feedback as part of the research process. Additional copies may be obtained

by contacting Jeane Jenkins at the SCRSJ’s Glasgow address below or may be downloaded

from our website.

Written comments on the report are welcomed. These should be sent to Nick Bailey at

the SCRSJ’s Glasgow address by Friday 6 June 2003. Attributed comments may be

included in the final report, unless respondents indicate that they wish their comments to be

treated in confidence.

Scottish Centre for Research on Social Justice (SCRSJ)

The Scottish Centre for Research on Social Justice is based jointly in the Department of

Urban Studies, University of Glasgow and the Arkleton Centre for Rural Development

Research, University of Aberdeen. The SCRSJ's mission is to promote better understanding

of, and more informed debate about, the nature of social justice in Scotland, particularly in

relation to public policy. It is funded for an initial three-year period by a Research

Development Grant from the Scottish Higher Education Funding Council. Further details are

available from our website.

SCRSJ SCRSJ Department of Urban Studies Arkleton Centre for Rural Development Research University of Glasgow University of Aberdeen 29 Bute Gardens Old Aberdeen Glasgow G12 8RS Aberdeen AB24 3UF Tel: 0141 330 2094 Tel: 01224 273901 Fax: 0141 330 2095 Fax: 01224 273902 Email: [email protected] http://www.scrsj.ac.uk/

mailto:[email protected]

http://www.scrsj.ac.uk/

Contents

Executive summary..............................................................................................i

1. Introduction.....................................................................................................1

1.1 About this report 1

1.2 Consultations and feedback 2

1.3 Structure of the report 4

2. Definitions and conceptual framework.........................................................5

2.1 Defining deprivation 5

2.2 Deprivation and poverty 7

2.3 Deprivation and social exclusion 9

2.4 Deprivation, social cohesion and social capital 11

2.5 Deprivation and social justice 13

2.6 Conclusions 14

3. Uses, context and approaches ......................................................................16

3.1 The uses of deprivation measures 16

3.2 The Scottish context 17 3.2.1 Geographical context 17 3.2.2 Cultural context 21 3.2.3 Institutional context 22 3.2.4 Conclusions 22

3.3 Individual and area approaches 23 3.3.1 Individual approach 23 3.3.2 Area-based approach 24 3.3.3 Conclusions 25

4. Developing area-based measures of deprivation .......................................26

4.1 Existing domains 27 4.1.1 Income deprivation 27 4.1.2 Employment deprivation 31 4.1.3 Health deprivation and disability 33 4.1.4 Education, skills and training deprivation 35 4.1.5 Geographical access to services 37

4.2 Other domains 40 4.2.1 Financial resources and financial exclusion 41 4.2.2 Housing deprivation 42 4.2.3 Physical environment 45 4.2.4 Crime and social order 46 4.2.5 Geographical access to employment 47 4.2.6 Communications and digital exclusion 49 4.2.7 Social relationships, social capital and social status 50

4.3 Units for analysis 51

4.4 Domains versus life-stage approaches 55

5. Developing measures of individual deprivation.........................................58

5.1 Measuring individual deprivation 58

5.2 Measures for different social groups 62 5.2.1 Children and young people 63 5.2.2 Older people 65 5.2.3 Women and men 66 5.2.4 Black and minority ethnic groups 67 5.2.5 People with disabilities 68 5.2.6 Conclusions 70

6. Updating the measures .................................................................................71

7. Conclusions ....................................................................................................72

References ..........................................................................................................73

List of figures and tables Figure 2.1: Poverty and deprivation – Townsend’s framework 8

Figure 2.2: Poverty and deprivation – Nolan and Whelan’s framework 9

Figure 4.1: Average ward population by Unitary Authority 53

Table 4.1: SIMD2003 indicators by domain and life-stage 57

Table 5.1: Necessities of life in Britain, 1999 60

Abbreviations

ID2000 – the Index of Deprivation 2000, compiled by the Oxford team for DETR, containing

six domain indicators and an overall index for English wards, as well as summary

measures for local authorities.

IMD2000 – the ward-level Index of Multiple Deprivation, i.e. a sub-set of the ID2000

PSE1999 – the Poverty and Social Exclusion Survey 1999, conducted by a team led by

Bristol University, and based on a survey of 1530 individuals across Britain.

SHS – the Scottish Household Survey. A continuous survey commissioned by the Scottish

Executive, with an annual sample of around 15,000 households. It is designed to

produce reliable estimates of local authority characteristics over a two year period by

sampling around 30,000 households in that time.

SIMD2003 – the Scottish Index of Multiple Deprivation, compiled by at team at Oxford

University led by Mike Noble. The index is at the level of wards in Scotland (as at

1999) using data for 2001, and was published in 2003.

MEASURING DEPRIVATION IN SCOTLAND – INTERIM REPORT – EXECUTIVE SUMMARY

Executive summary

1. Introduction

• This is the first report from a project to produce a long-term strategy for measuring

deprivation in Scotland. It is an Interim Report which has been published to encourage

wide discussion of the issues and feedback. The project aims:

o to provide a clear definition and conceptual basis for measuring deprivation;

o to set out a long-term strategy for measuring area deprivation; and

o to explore approaches to measuring deprivation for individuals and for different

social groups.

• The project will work from the definition and conceptual framework to identify the

indicators and data sources necessary to provide proper measures of deprivation in

Scotland. We will make recommendations where additional data sources should be

developed. The work will not be limited by considerations of what data is currently

available.

2. Definitions and conceptual framework

• This section sets out our definition of deprivation and explores how it relates to others

used to describe social needs, such as poverty, social exclusion and social justice.

• Deprivation is valuable for its focus on the enforced lack of goods, services or social

relations which result from a lack of financial resources. It is a relative measure where

standards are defined in relation to social norms or expectations.

• Poverty and deprivation are therefore interlinked as cause and outcome. Both financial

resources and outcomes should be captured in measures of multiple deprivation.

• We see deprivation and social exclusion as closely related concepts while recognising a

diversity of views within the exclusion literature in particular. Both are multi-dimensional

concepts where issues of duration are relevant. Social or relational dimensions are

important to both, although they have been given greater emphasis in the exclusion

literature. While some commentators do not consider poverty to be an essential feature of

social exclusion, deprivation measures can provide direct measures of some forms of

exclusion and indirect evidence of other forms.

• Social cohesion draws attention to a range of issues including social connectedness or

social capital and social status. The discussion of social capital issues raises questions

about the types of social relationship which deprivation measures should focus on and

i


about the relationship between income and social connectedness. The discussion of social

status issues raises questions about the extent to which people on low incomes feel

marginalised or stigmatised by others, and whether such issues should be captured in a

deprivation measure.

• There is a divide within the social justice literature between those who focus on

individual material equalities and those who emphasise group identity and cultural

equality. To some extent, this mirrors differences within the social exclusion literature.

As with exclusion, deprivation measures may provide good measures of certain injustices

and starting points for understanding others. Comparisons of levels of deprivation

between different social groups may provide important evidence of the existence of

cultural inequality or discriminatory processes. The ability to disaggregate deprivation

measures for relevant social groups is therefore important.

3. Uses, context and approaches

• Deprivation measures might be used to provide a broad picture of relative levels of needs

(between areas or social groups) to guide the allocation of resources or to monitor change

over time. Alternatively, measures might emphasise depth of coverage and better

understanding of the factors causing deprivation to guide policy design. Our remit is

primarily related to the first of these although we recognise the benefit of deeper

approaches in the development of broad measures.

• For the purposes of measuring deprivation, differences between Scotland and the rest of

the UK are limited. Approaches adopted elsewhere are likely to be useful in the Scottish

context. The most significant area of difference is in the extent to which Scotland has a

larger population living in very remote areas. Differences in physical geography, culture

or institutional contexts appear relatively minor.

• Both individual and area-based approaches to measuring deprivation may be useful as

they provide different measures for different purposes. Interviewees tended to express a

similar view. Our initial recommendation would therefore be to develop both.

ii


4. Developing area-based measures

• The first part describes the current index, and analyses both the conceptual basis of each

domain and the specific indicators used.

• Section 4.2 discusses possible additions to the list of domains. Our aim is to identify

significant aspects of deprivation not yet captured by the index. We cover: financial

resources beyond current income; housing; physical environment; crime and social order;

geographical access to employment; communications and social relationships and social

capital. We outline the scope of each potential domain, its conceptual basis and the

measures which might be adopted.

• In 4.3, we discuss the different spatial units which might be used. The move to wards has

been widely welcomed. There was a desire to see even smaller units used but recognition

that this would mean fewer indicators. We note that wards provide a finer level of detail

in more rural authorities which may be considered fair. We discuss the “small data units”

developed for the Neighbourhood Statistics project.

• In 4.4, we provide some criticisms of the current domain-based approach and set out an

alternative (or complementary) approach based on life-stage groups.

5. Developing individual measures

• There was strong support for developing a measure of individual deprivation, particularly

one which would provide estimates down to the level of local authorities. This would act

as a complement to area-based measures. We see the approach adopted in the PSE1999 as

providing the best template for this work as it relies on direct measures of living

standards, tested against a standard determined by a consensual approach.

• In some cases, there is an argument for developing a separate standard to capture levels of

need within a particular social group more accurately. On the basis of our initial analysis,

the clearest case would appear to exist in relation to children, women and disabled people.

The arguments for older people, young people and people from black and minority ethnic

groups are less clear cut. We do not discuss different sexuality groups but would be keen

to comments on their relative needs.

iii


6. Updating the measures

• There is a trade-off between retaining a consistent measure to enable analyses to be made

of change over time and updating a measure to take account of new data sources,

methodological advances or changes in social perceptions or practices. With the move to

produce deprivation measures on a more regular basis, possibly annually, a compromise

position might be useful. This would retain the same measure for a period of several

years. Several revisions could then be made at one time rather than minor changes made

each year. By ensuring overlap in the change years, a more continuous picture of the

changing pattern of deprivation could be developed.

• With area-based measures, area boundaries would need to remain consistent to enable

comparisons to be made over time. Again, updating could be carried out periodically

although it would probably not be possible to produce data for two different sets of

boundaries in the same year due to problems of disclosure.

7. Conclusions

• We do not claim that deprivation is a better measure of social needs than any other

concept. No one concept will meet the needs of all situations or potential users. We do

believe, however, that deprivation is important for its focus on the impacts of poverty or

lack of financial resources on material and social living standards.

• In discussing how to measure deprivation, we have tried to keep to these core issues and

not to let the concept expand so far that its meaning is lost. Nevertheless, the overlap

between terms such as deprivation and social exclusion or social justice means that the

measurement of deprivation can provide important insights into a wider set of problems.

iv

MEASURING DEPRIVATION IN SCOTLAND – INTERIM REPORT

1. Introduction

1.1 About this report

This is the first report from a project to produce a long-term strategy for measuring

deprivation in Scotland. It is an Interim Report which has been published to encourage wide

discussion of the issues and feedback. The project has been commissioned by the Central

Statistics Unit of the Scottish Executive as part of their programme of work on developing a

Neighbourhood Statistics database for Scotland. It comprises three main elements:

• The project will provide a clear definition and conceptual basis for measuring

deprivation. In particular, it will clarify how the term relates to others used to refer to

social needs such as poverty, social exclusion or social injustice.

• The project will set out a long-term strategy for measuring area deprivation. It will

ensure that the measure gives full coverage to deprivation in the Scottish context, and

to different contexts within Scotland, particularly rural and urban.

• The project will explore approaches to measuring deprivation for individuals. It will

recommend whether individual measures should be developed as replacements to

area-based measures or as additional to them. It will also explore the desirability and

feasibility of developing a range of measures related to the specific needs of different

social groups.

The aim of the project is to work from the definition and conceptual framework to identify

the indicators and data sources necessary to provide proper measures of deprivation in

Scotland at area and individual level. In particular, we have been asked to make

recommendations where additional data sources should be developed. The work will not be

limited by considerations of what data is currently available although issues of cost, value-

for-money and practicality will of course be taken into account.

The second element builds on the recently published Scottish Index of Multiple Deprivation

2003 or SIMD2003 (Noble et al, 2003). The team from Oxford University was commissioned

to construct an area-based index of deprivation using essentially the same method as had been

1


used in England, Wales and Northern Ireland. Data was obtained for all 1222 wards in

Scotland (using 1999 boundaries). The SIMD2003 represents a considerable advance on

previous Scottish area deprivation indices.

The aim of this Interim Report is to encourage feedback from a wide range of groups with an

interest in the measurement of deprivation. It sets out our initial thinking on each of the three

main research elements. In some areas, our work is relatively well developed and we set out

initial proposals. In other areas, our work is at an earlier stage and we are looking for

comments or suggestions to take things forward. Throughout the report, we try to be specific

about the questions on which we would like comments. Direct questions are there to help

structure responses, however, not to limit them. Respondents are invited to make any

comments on matters related to this work.

The final deadline for written comments is Friday 6 June. Comments should be sent to

the research team at the address given inside the front cover of this report. Attributed

comments may be included in the final report, unless respondents indicate that they wish their

comments to be treated in confidence. Although we will take account of feedback from the

consultation process, the final recommendations which emerge from this work will be those

of the research team. The Final Report will be clear, however, where our recommendations

have received broad support from those consulted and where there has been opposition or

disagreement.

1.2 Consultations and feedback

This Interim Report has been produced following an initial period of research, which

included:

• a review of relevant literature on the definition, conceptualisation and measurement of

deprivation and related terms;

• interviews with around 20 representatives of central government, local government, other

public agencies, the voluntary sector and academics;

• two focus groups conducted with representatives of community organisations from a

variety of rural and urban locations; and

2


• a number of written submissions from representatives of public and voluntary groups,

many based on the interview schedule which was published on the SCRSJ website.

The initial consultations were intended to provide a range of views and perspectives on the

issues raised by this project. They were not intended to be representative of any one group or

to cover all the views within each group. Rather, they acted as a preliminary element in

broadening the issues identified by the research team and in grounding the work in the

experiences, needs and perceptions of a range of interest groups.

Publication of this report marks the start of a more formal consultation stage. In addition to

providing the report, we will be holding three public meetings in May. The purpose of these

meetings is to publicise the work, to provide an opportunity for discussion of the issues and

our initial recommendations, and to get feedback on the work. Given the scale and

complexity of the issues, these meetings cannot hope to cover every aspect of the work and

we hope that many of those attending will also make more detailed written comments. Details

of the meetings are provided below.

To assist us with planning, it would be very helpful if people could let us know if they

are planning to attend a particular meeting by contacting Jeane Jenkins

([email protected] or 0141 330 2094). Directions to each venue can also be obtained

from Jeane Jenkins.

Location Date and Time Venue

Edinburgh Tues 20 May 10am – 12pm (9.30am Tea/Coffee)

Raeburn Room SCVO Offices Mansfield Traquair Building 15 Mansfield Place

Inverness Tues 27 May 2pm – 4pm (1.30pm Tea/Coffee)

Netley Centre Bishop’s Road

Glasgow Wed 28 May 2pm – 4pm (1.30pm Tea/Coffee)

Glendornoch Boardroom Institute of Electrical Engineers Teachers Building St Enoch Place

3


1.3 Structure of the report

The structure of the report is as follows:

• Section 2 covers the first element of our work – providing a definition and conceptual

framework for measuring deprivation, and setting out how we see deprivation in

relation to other terms used to refer to social needs.

• Section 3 deals with some general issues which relate to the measurement of

deprivation. It discusses different uses for deprivation measures, before setting out the

distinction between area-based and individual measures. This section also looks at the

Scottish context and how this might influence measures.

• Section 4 turns to the second main element of this work – developing a long-term

strategy for measuring area deprivation. It starts with a detailed consideration of the

indicators and domains currently included in the SIMD2003, before discussing

possible new domains. This section also examines the most appropriate spatial units

for the index as well as considering whether the current approach based on “domains”

could be supplemented or replaced by one based on “life-stages”.

• Section 5 covers the third main element – individual and social group measures of

deprivation. The first part looks in general terms at how multiply deprived individuals

might be identified using a single standard. The second part presents a preliminary

exploration of whether specific measures should be developed for different social

groups and how these might differ from the single standard.

• Section 6 briefly sets out some issues to be considered in updating deprivation

measures (at area or individual level) to take account of the development of new data

sources or methods.

• Section 7 provides a brief conclusion to this stage of the work.

4


2. Definitions and conceptual framework

People use terms like deprivation in different ways. No one definition would be agreed on by

everyone. Nevertheless, it is necessary for us to provide a single definition on which to base

the rest of our work. We also need to provide definitions of other terms such as poverty,

exclusion and injustice in order to describe the relationships between deprivation and these

terms. This also enables us to see the extent to which a measure of deprivation is a measure

of these other social phenomena at the same time.

Some people see all these terms as very similar: “all are linked” and “we deal with these

words every day,” as two community representatives said during the initial consultation

process. Readers who take this view may prefer to read only the conclusions for this section

before moving to the remaining sections which deal with measurement. We see this as a

necessary stage if we are to be clear about what we are trying to measure – and what we are

not. Other people see important distinctions between the terms and favour one over the other:

“poverty is the best term, better than deprivation” or “social capital should be at the top”.

These comments from two focus group respondents are mirrored by similar views among

some researchers and policy makers. We are not arguing that deprivation is a “better” term

than any other or that it should be used in place of other terms. Rather our concern is to

provide a clear definition of one term, deprivation, which we consider to be useful and

relevant to the analysis of some social problems in contemporary Scotland.

2.1 Defining deprivation

In common with most researchers, our starting point is the work of Peter Townsend:

“Deprivation takes many different forms in every known society. People can be said

to be deprived if they lack the types of diet, clothing, housing, household facilities and

fuel and environmental, educational, working and social conditions, activities and

facilities which are customary, or at least widely encouraged and approved, in the

societies to which they belong.” (1987: p.126)

5


There are five elements to this definition. First, it focuses on living standards and, in

particular, the extent to which people lack specific items such as clothing, housing and so on.

Second, it is multi-dimensional, as people can be deprived in different ways. There is

consequently a requirement to measure deprivations across a range of different dimensions or

domains. Each item that a person lacks may be seen as a separate deprivation and people

lacking a given item may be termed “deprived” in that respect. It should be stressed,

however, that the term is usually used to refer to people who have several deprivations and

who are therefore suffering from “multiple deprivation”. At times, the term “deprivation” is

used as shorthand for “multiple deprivation” but it is the latter which is our focus.

Third, Townsend’s definition is concerned with both material and social or relational

dimensions. The latter refers to the ability of an individual to participate in the normal social

life of their community – visiting family, having friends round, or attending social events

such as birthdays, weddings or funerals, for example. Townsend and others have noted that

most attempts at measurement have focused on material dimensions, as these are easiest to

capture. However, he identifies social aspects as essential to our understanding of the nature

and impacts of deprivation. Many of the interviewees and focus group participants involved

in our initial consultation process agreed with this perspective.

Fourth, Townsend’s definition is a relative one. It is based on socially accepted norms or

standards which will differ from one society to the next, and which will change over time.

There is a need to check that any measure does indeed reflect widely held views about what

are minimum acceptable standards of living and that it is updated in line with changing public

perceptions. Fifth, Townsend’s definition focuses on individuals not areas. Although some

aspects of the social, economic and physical environment in which people live can be

important factors in deprivation, it is people who are deprived, not areas.

Townsend’s approach has been refined further, notably in the work on the Breadline Britain

surveys and the Poverty and Social Exclusion 1999 survey (Mack and Lansley, 1985; Gordon

and Pantazis, 1997; Gordon, Adelman et al, 2000). These have provided a more systematic

method for defining which “necessities” of life people in a given society should expect to

have. This is done by consensus; necessities are defined as those items which at least 50 per

cent of the population believe an individual needs in order to participate in everyday life.

These surveys also attempt to distinguish between people who lack an item through choice or

6


preference, and those who lack it due to inadequate income or resources, with only the latter

being identified as deprived.

One criticism of the consensual approach is that it can be seen as backward looking. In

general, surveys have shown a gradual expansion in the list of goods regarded as necessities,

reflecting rises in living standards (Gordon, Pantazis and Townsend, 2000). A view expressed

in the interviews was that we should be more forward looking, including those items now that

are likely to play an increasingly important role in future. A current example would be access

to a computer or to the Internet from home. The main counter-argument is that this approach

lacks the direct legitimacy of the consensual approach. Who gets to decide which items will

be necessities in future and what criteria should be used? It is particularly difficult to argue

that something is a necessity when less than half the population currently has it.

Many writers draw a sharp distinction between deprivation and social status which is defined

by factors such as age, ethnicity or household type. Some status groups experience much

higher levels of deprivation than others on average. In Scotland in recent years, deprivation

levels have been higher among children, older people, some minority ethnic groups and lone

parent households, for example. However, membership of one of these groups is not a cause

of deprivation. “Even if many such people are deprived, it is their deprivation and not their

status which has to be measured. And many people having that status are demonstrably not

deprived” (Townsend, 1987, p. 135). This does not mean that status is irrelevant to

discussions of deprivation. In this report, however, status is treated as a risk factor, not as a

deprivation in itself.

2.2 Deprivation and poverty

Many accounts of deprivation are very explicit about the relationship between poverty and

(multiple) deprivation. For Townsend, deprivation is the direct result of poverty (lack of

income or, more generally, lack of financial resources). He argues that the relationship

between poverty and deprivation is so strong that it is possible to identify those living in

deprivation either through direct measures of standards of living (deprivation measures) or

through indirect measures of incomes or resources (poverty measures). This is illustrated in

Figure 2.1. He acknowledges that some individuals may suffer from deprivation without

7


being in poverty and that others in poverty may not suffer deprivation, but sees this as a

minor issue.

Figure 2.1: Poverty and deprivation – Townsend’s framework

Poverty as lack of financial

resources (relative to needs)

measured by low income

LEADING

TO

Deprivation as lack of

necessities, both material and

social/relational

Townsend also argues that there is a particular threshold in the income distribution where

further decreases in income bring about sharp rises in levels of deprivation: “that ‘threshold’

properly marks the beginning of a state of objective poverty” (Townsend, 1987: p. 130).

Townsend uses his survey evidence to try to identify the threshold point more precisely.

Several studies have argued that the relationship between income and deprivation is not as

clear cut as Townsend suggests (Piachaud, 1987; Ringen, 1988). First, there are questions as

to whether there is a distinct threshold or tipping point at all or whether the relationship is

more continuous. Second, there are questions about the strength of the relationship. Some

evidence suggests that, for any given level of income, levels of deprivation vary quite widely.

Not everyone below the official poverty line is deprived. A key explanation for this is that the

financial resources which households have are not only determined by current income.

This insight is central to work which takes a broader approach to measuring the level of

resources available to an individual. For Nolan and Whelan (1996), resources reflect not only

current income, but also levels of savings or debts. This approach highlights the importance

of the duration of poverty spells in determining deprivation outcomes. For example, when an

individual initially faces a drop in income, they may be able to avoid a fall in living standards

(and hence deprivation) by drawing on savings but, as time goes on, they are likely to face a

decline in their living standards as savings are depleted. There is also a link with past

employment history, as frequent periods of unemployment limit the accumulation of savings

and increase the likelihood of accruing debts. Poverty in retirement is strongly linked to

previous work history. Social relationships also become important as an individual’s level of

8


financial resources may depend on their ability to draw on financial support from family and

friends in times of need.

Nolan and Whelan (1996) also argue that identifying individuals in deprivation requires

measures of both resources and outcomes or living standards to ensure that the measure

captures individuals with low standards of living resulting from lack of resources. The

relationship is illustrated in Figure 2.2. The SIMD2003, like previous area deprivation

indices, includes measures of both resources (incomes) and living standards (outcomes) and

in this sense appears closer to this conception than to Townsend’s. We also favour this

refinement of Townsend’s model and will seek to use it as the basis for our work.

Figure 2.2: Poverty and deprivation – Nolan and Whelan’s framework

Multiple deprivation caused by lack of resources:

Poverty as lack of financial

resources (relative to needs)

measured by low income but also

other factors such as: savings or

debts; family or friends who can

help; stability of income; social

background; work history

AND

Deprivation as lack of

necessities due to lack of

resources. Necessities are both

material and social/relational.

2.3 Deprivation and social exclusion

The Social Exclusion Unit’s definition of social exclusion is often quoted in public policy

discussions. This sees exclusion as “what can happen when people or areas suffer from a

combination of linked problems such as unemployment, poor skills, low incomes, poor

housing, high crime environments, bad health and family breakdown” (SEU website). This

resonates with the domains included in many indices of area deprivation such as the

SIMD2003. Most accounts of social exclusion, however, define it in ways that emphasise its

applicability to individuals rather than areas.

9


Two of the characteristics that most conceptualisations have in common are the multi-

dimensional and dynamic nature of exclusion (Room, 1995). In this respect, the differences

between exclusion and deprivation are ones of emphasis. While both include material and

social or relational dimensions, exclusion studies tend to give greater weight to the latter. As

already noted, many deprivation measures have relied heavily on material aspects as these are

most easily measured; some are referred to as indices of “material deprivation” to make the

distinction clear (Carstairs and Morris, 1991). This approach would not generally be

considered sufficient for measuring exclusion. We are aware of this weakness in deprivation

studies and see it as an important gap to be filled.

The focus on the dynamics of exclusion draws attention to the duration of situations

experienced by individuals. Being on a low income only briefly can have quite different

implications for individual living standards compared with a prolonged spell in poverty.

While this has been under-emphasised in earlier work on deprivation, it is certainly present in

some approaches as the discussion of Nolan and Whelan’s (1996) work showed. We would

seek to incorporate this insight into our work on deprivation.

Many accounts of social exclusion focus on the processes which lead to (or fail to prevent)

exclusion rather than the outcomes of these processes. Shucksmith and Philip (2000), for

example, provide a typology of four such systems: private, representing market processes;

state, representing public processes; voluntary; and personal (family and friends). For authors

such as these, an understanding of social exclusion requires processes to be studied, using

qualitative methods. The focus is on understanding social exclusion rather than on measuring

it. ‘Outcome’ measures of deprivation provide no more than a clue to the existence of

exclusionary processes.

There are a number of divergences between social exclusion and deprivation literatures

(Berghman, 1995; Levitas, 1998; Hills et al, 2002). We highlight just two here. First, some

people use the term social exclusion in preference to deprivation because they see poverty or

material living standards as just one dimension of a wider set of problems. They argue that

social exclusion may arise where people are prevented from participating in economic, social,

cultural or political aspects of society, regardless of whether they are in poverty or not. To be

included requires not just adequate income (absence of poverty) but also access to work for

those who want it, social and cultural integration, and political empowerment. From this

perspective, the concept of deprivation appears more narrowly focussed. Other writers on 10


social exclusion argue that it is difficult, if not impossible, to separate social, cultural or

political marginalisation from issues of poverty or deprivation; the most marginalised tend to

be the poorest. From this point of view, the gap between deprivation and social exclusion

concepts should not be overstated. In either case, deprivation measures provide direct

evidence on some aspects of social exclusion and indirect evidence of other aspects.

Second, some views of social exclusion see it as the process of becoming detached from a

“moral order” that supposedly binds a society in a set of mutual rights and obligations

(Room, 1995, p. 6; Etzioni, 1995). Within deprivation studies, the focus on social deprivation

has been limited to the ability to maintain social relationships without any concern for

individual values. We would not intend changing this approach.

2.4 Deprivation, social cohesion and social capital

Social cohesion and social capital are usually presented as features of a good society or

community. From a policy perspective, they are often seen as absent, reduced or under threat

amongst deprived or excluded groups. Moreover, this can be held to the cause of a number of

social problems so that enhancing social cohesion or social capital is seen as a priority (for

example, Scottish Executive, 2002a). It should be noted that both views may be disputed. In

our initial consultations, however, there was considerable encouragement to explore the lack

of social cohesion or capital as possible dimensions of deprivation.

As with social exclusion, people use the concepts of social cohesion or capital to draw

attention to different issues. Buck et al (2002) provide one useful typology, recognising three

sets of concerns: social connectedness or social capital; social equality or status; and social

order.

An individual’s social connectedness or social capital refers to the set of personal

relationships or networks which they have by virtue of their contacts with family, friends,

neighbours, colleagues or other acquaintances. Different types of social capital may be said to

perform different functions or bring different benefits. One of the best known distinctions is

between “bonding” and “bridging” social capital, which builds on Granovetter’s (1973)

distinction between “weak” and “strong” ties. Strong ties exist between family and close

friends and these are seen as valuable in helping a person “get by” (bonding capital). They 11


have limited value in helping you “get on” (bridging capital), however, as the “best friends”

of your “best friends” are likely to be people you know or come into contact with in any case.

Weak ties on the other hand tend to exist between looser acquaintances such as work

colleagues. While these do not tend to be a source of support, they may be more valuable as

sources of information or further contacts because they connect you to much wider circles of

people. There is much less overlap in the social networks of acquaintances connected only by

weak ties. People living in areas with high concentrations of deprived households are

sometimes said to suffer from “network poverty” as they have high levels of strong ties but

relatively few weak ties.

There is some degree of overlap between the concepts of social capital and deprivation, as the

latter is concerned with an individual’s ability to maintain or develop social relationships as

well as with material living standards. While deprivation appears to be concerned primarily

with the maintenance of social relationships as a “necessity of life” or as an end in itself,

however, the social capital literature is concerned with these relationships as means to other

ends. More importantly, deprivation studies have tended to emphasise closer personal

relationships rather than the wider social networks which some conceptions of social capital

see as important. In this respect, there appears to be some tension between the two

perspectives. In addition, deprivation theories would suggest that the ability to maintain or

develop social relationships is restricted by low income, while some empirical studies of

social capital suggest that certain social relationships (strong ties) are more likely to be

present for people on lower incomes. Two areas for further research and discussion are the

types of social connection which deprivation studies should focus on and the relationship

between income and social connectedness. These are issues we return to below when

considering appropriate measures of social deprivation in Section 4.2.7.

The second focus of social cohesion studies, social equality, reflects a concern with status

differences between individuals. One question for deprivation studies is the extent to which

people on low incomes feel that have a low social status or feel marginalised or stigmatised

because of their material position. Possible approaches to measuring the (lack of) social

equality or status at the individual level might focus on factors such as the degree to which a

sense of stigma or low status is felt, the degree to which people feel respected as equal to

others and so on. Such measures would be innovative and therefore care needs to be taken in

developing them if they are to have widespread support. If operationalised successfully, they

could be particularly valuable for monitoring equalities policies. 12


The third focus within social cohesion is on problems of social order. These are strongly

linked to problems of crime and social disorder often seen as characteristic of areas with high

concentrations of deprivation (SEU, 1998). The implication is that, to some extent, the

weakness of communities in these areas enables or permits criminal or disorderly activities to

occur – a “blame-the-victim” approach. This is not to deny that crime or fear of crime may be

important issues for many people on low incomes, but we would not tend to see these in

terms of problems of social relations or community failure.

2.5 Deprivation and social justice

Social justice is generally accepted as being concerned with “how the good and bad things in

life should be distributed” (Miller, 1999, p. 1). A few accounts (Nozick, 1974, for example)

restrict their concern to processes of distribution and to ensuring “procedural justice” or fair

treatment in the different distributional systems within society, including both market and

state systems. For the great majority, however, the concern is with achieving fair or just

outcomes, and the fairness of different processes needs to be judged in relation to these

outcomes.

As with debates on social exclusion, there is a divide within the social justice literature

between conceptions which emphasise material distributions and those which emphasise

social status or identity (Phillips, 2000). The first group has much in common with the

deprivation literature and with material conceptions of exclusion. They are concerned with

individual living standards although there are many differences within the group, in

particular, between those which emphasise equality of outcomes and those which emphasise

equality of opportunity. For the former, measures which demonstrated unequal outcomes

would tend to be seen as evidence of injustice. For the latter, there would need to be further

evidence that individuals had made equal efforts to take advantage of opportunities available

to them (to seek work, for example) before unequal outcomes would be seen as unjust.

Deprivation measures may therefore provide good measures of injustice if justice is seen in

terms of equality of outcomes. If it is seen as equality of opportunity, then deprivation

measures may provide good starting points for further investigation of the processes

involved.

13


The second group within social justice has focussed on social groups and issues of status. For

these conceptions, the key issue is equality of status, including recognition of identity, and

equality of access and respect in political processes. There is an emphasis on the need for a

just society to respect the different identities or cultures which groups have as well as

ensuring material equality between them (Young, 1990). From this perspective, measures of

material inequalities or deprivation might be seen as providing only limited measures of

injustice. As we argued above in relation to social exclusion, however, there are likely to be

important relationships between social or cultural marginalisation and material position. In

this respect, differences in levels of deprivation between different social groups may provide

important insights into the existence of social or cultural injustices. The ability to

disaggregate deprivation measures by social group would therefore be highly valued from

this perspective.

2.6 Conclusions

The aim of this section was to set out our understanding of what deprivation is and how it

overlaps with, or differs from, other terms used to refer to social needs. None of the terms has

a single, uncontested meaning although deprivation emerges as one on which there is

relatively close agreement.

• Deprivation is valuable for its focus on the enforced lack of goods, services or social

relations which result from a lack of financial resources. It is a relative measure where

standards are defined in relation to social norms or expectations.

• Poverty and deprivation are therefore interlinked as cause and outcome. Both financial

resources and outcomes should be captured in measures of multiple deprivation.

• We see deprivation and social exclusion as closely related concepts while recognising a

diversity of views within the exclusion literature in particular. Both are multi-dimensional

concepts where issues of duration are relevant. Social or relational dimensions are

important to both, although they have been given greater emphasis in the exclusion

literature. While some commentators do not consider poverty to be an essential feature of

social exclusion, deprivation measures can provide direct measures of some forms of

exclusion and indirect evidence of other forms.

14


• Social cohesion draws attention to a range of issues including social connectedness or

social capital, social status and social order. The discussion of social capital issues raises

questions about the types of social relationship which deprivation measures should focus

on and about the relationship between income and social connectedness. The discussion

of social status issues raises questions about the extent to which people on low incomes

feel marginalised or stigmatised by others, and whether such issues should be captured in

a deprivation measure.

• There is a divide within the social justice literature between those who focus on

individual material equalities and those who emphasise group identity and cultural

equality. To some extent, this mirrors differences within the social exclusion literature.

As with exclusion, deprivation measures may provide good measures of certain injustices

and starting points for understanding others. Comparisons of levels of deprivation

between different social groups may provide important evidence of the existence of

cultural inequality or discriminatory processes. The ability to disaggregate deprivation

measures for relevant social groups is therefore important.

Views are invited on the definition and conceptualisation of deprivation presented

above, and on the analysis of how deprivation relates to the other terms discussed.

15


3. Uses, context and approaches

This section covers a number of general issues relating to the measurement of deprivation. It

starts by discussing the different purposes which may underpin measurement exercises. It

then examines the distinctive features of Scotland that may influence the nature of

deprivation here, and hence the design of deprivation measures. Finally, it outlines the basic

differences between measuring deprivation at an individual and an area level as a prelude to

the more detailed discussions in Sections 4 and 5 below.

3.1 The uses of deprivation measures

As well as knowing what is to be measured, it is important to know why it is being measured

when making decisions about the design of a measure. Deprivation measures may have a

wide range of uses, as our initial consultations revealed.

Many interviewees stressed the value of these measures as means of assessing levels of

needs. Relative measures would be of value in identifying priority areas or groups or guiding

resource allocations. There is a long history of using area deprivation indices in this way.

These are relative measures in the sense that they rank areas in relation to each other but

provide no indication of the number of people in each area in (multiple) deprivation. An area

may improve its ranking from one year to the next without necessarily seeing any reduction

in the proportion of people deprived.

Increasingly, there is also interest in absolute measures to monitor the extent to which levels

of deprivation change over time. These might be of value in evaluating the effectiveness of

different policies or interventions or for judging value-for-money. Community Planning

Partnerships are now responsible for “closing the gap” between the most deprived areas and

the average (Scottish Executive, 2002a) and this implies the use of precisely this sort of

absolute measure. Both types of use stress the need for broad coverage of the population

using a single standard to measure the extent of deprivation and to make comparisons

between different groups or areas.

16


Some interviewees also argued that there was a need for measures which provided greater

detail on the nature and causes of deprivation. These would need research to cover not only

the core indicators of deprivation but also a much wider set of information on individuals’

characteristics, backgrounds and circumstances to explore the factors associated with

deprivation. Here the emphasis was on informing the design of policies or interventions. This

implies depth rather than breadth of coverage.

Given the wide range of potential uses for deprivation measures, no one measure could hope

to meet all needs. Our remit was framed more clearly in terms of the first of these two

concerns and this is the type of measure we focus on. There is of course a close relationship

between the two. As knowledge and understanding of the causes of deprivation improve, this

would feedback into broader measures.

Views are invited on this summary of the uses of deprivation measures and, in

particular, on the value of producing measures primarily to assess levels of need.

3.2 The Scottish context

The basic processes of economic restructuring and social change which have driven rising

levels of poverty, deprivation and social exclusion in Scotland have been the same as those

operating in the rest of the UK. The situation in Scotland is not fundamentally different to

other parts of the country. However, there are distinctive features of the geographical, cultural

and institutional context which may suggest a need for a Scottish variation or Scottish

conceptualisation of deprivation (Brown et al., 2002).

3.2.1 Geographical context

Physical environment

There are some climatic differences between Scotland and other parts of the UK which may

have a bearing on needs. Scotland’s colder, wetter and windier climate, particularly on the

west coast, is likely to increase the deprivation problems associated with inadequate housing,

notably fuel poverty arising from inadequate heating systems and insulation. Although not

17


unique to Scotland, it becomes more important to ensure that these problems are adequately

captured.

Urban and rural dimensions

A more significant issue is the challenge of providing measures of deprivation which cope

with the range of situations in which people live in Scotland – from dense neighbourhoods in

large metropolitan areas to small settlements or isolated homes in remote and sparsely

populated regions. There is great diversity between these extremes as well and the terms

‘urban’ and ‘rural’ do not reflect this adequately. While we continue to use these terms as a

convenient shorthand, we need to bear this diversity in mind at all times.

Recent evidence shows that, on a wide range of indicators, urban areas as a whole tend to

have higher levels of deprivation than rural areas. In Scotland, the Scottish Executive’s Social

Justice milestones report for 2002 provided an urban/rural breakdown for 12 of the

milestones, covering education, health, employment, crime, volunteering and digital

exclusion (Scottish Executive, 2002b). While the strength of the relationship varied, urban

areas scored worse than rural on 11 of these. The one exception was access to the internet

where there was no difference between urban and rural areas. In England, Harrop and Palmer

(2002) compared urban and rural areas on 50 indicators of poverty and social exclusion,

finding that urban areas had higher levels of need on 30, while rural areas had higher levels of

need on 3. For Britain as a whole, Chapman et al (1998) found a lower incidence of low

income and shorter spells on low income in rural areas, with the proportion “persistently

poor” significantly less although a similar study by Kempson and White (2001) found no

difference.

Importantly, all these studies show that there are significant levels of deprivation across the

country. Deprivation is not confined to urban areas. There is also a great deal of diversity

within both urban and rural categories, as well as some evidence of systematic differences; on

many indicators, people in remote rural areas are more likely to suffer deprivation than those

in accessible rural areas (Harrop and Palmer, 2002; Scottish Executive, 2002b).

The importance of the urban-rural issue is increased because, in the past, many area-based

deprivation indices were biased against rural areas due to the indicators used (Midwinter and

Monaghan, 1991; Shucksmith, 1990; Shucksmith et al., 1994, 1996). The most obvious

example has been the use of car-ownership as a proxy for income, but there are problems also 18


with other indicators such as the proportion of housing in local authority ownership or the

proportion of children in high flats. The use of these indicators reflected the fact that these

indices used to be constructed from data from the Census which was not designed to measure

deprivation. Failure to include indicators to reflect problems which are of greater importance

in rural areas (notably problems of remoteness and access) further compounded the problem.

Many of these criticisms have become redundant with the latest indices as these are based on

better data and different methodologies (Monk et al, 2003). Furthermore, the SIMD2003 and

its equivalents elsewhere in the UK have incorporated a new set of indicators on access to

services specifically to capture a major element of rural deprivation.

There are still some concerns that the new approaches may not fully reflect the nature of

problems in rural areas (Countryside Agency, 2003). Similar comments were made in our

interviews and focus groups. It is equally true, however, that there are aspects of urban

deprivation not captured by the current approach as we discuss in Section 4. The causes of

deprivation are complex whether in rural or urban areas. No attempt to summarise levels of

deprivation across wide areas will ever capture every detail of this and measures for small

areas will always be particularly limited by data availability. The important point is that these

measures capture the main dimensions of deprivation and that they do so in a balanced

manner in both rural and urban areas.

A further criticism of past deprivation indices from the rural perspective has been their focus

on concentrations of deprivation at small-area level. This is seen as disadvantaging rural

areas, as deprived households in these locations are more widely dispersed. Even if the

proportion of people deprived was similar in urban and rural areas as a whole, measures of

small-area concentrations would highlight the parts of urban areas where large numbers of

deprived individuals are concentrated. The problem with this criticism is that it confuses the

issue of measurement with the issue of use. Area-based measures do provide an accurate

picture of the concentration of deprivation in urban and rural areas even if they are seen as

being less useful in the latter case as deprivation is less concentrated there. In urban areas,

too, concern is expressed about the extent of hidden deprivation in “non-deprived” areas.

More positively from both rural and urban perspectives, this work will produce

recommendations on measures of deprivation both at the small-area level and at the

individual level. The latter will provide measures of levels of deprivation for Scotland as a

whole and for larger areas within it (possibly down to local authority level). It may be that the

latter are generally seen as more relevant in rural areas. 19


In some cases, criticisms of area-based deprivation measures have led organisations in rural

areas to work with different conceptions of social problems and hence different indicators to

identify areas of need. For example, Highlands and Islands Enterprise (HIE) and Highland

Council use the term “fragile areas” to refer to places facing long term problems of structural

economic change and decline. Others have used the broader term “disadvantage”

(Countryside Agency, 2003; Monk et al, 2003). These studies may provide valuable analyses

of rural social and economic problems but, from the perspective of this work, they do not

necessarily help with the identification of rural deprivation. While some of the specific

indicators for fragile areas or rural disadvantage could be deprivation measures (low

economic activity rates or high unemployment), others measure the probability of an area

facing problems in the future rather than current levels of need (elderly population; low rates

of in-migration of economically active people; dependency on primary sector employment,

for example). These are risk factors rather than deprivations in themselves.

Our starting point is that the basic dimensions of deprivation are the same across Scotland

(Rural Poverty and Inclusion Working Group, 2001). The core issues such as low income,

unemployment or poor quality of employment, poor health, poor housing or lack of access to

housing, poor education, poor services or lack of access to services affect people in rural and

urban areas, and are at the heart of what it means to be deprived in all cases. Having said this,

there are some differences in the manifestations and effects of deprivation and these are

reflected in the detailed discussions below. Interviews with representatives with a rural

interest suggested a number of additional dimensions of deprivation for consideration,

including access to a range of services as well as problems of social isolation. Issues of

remoteness and super-scarcity might also be considered, as Scotland has the only really

remote areas in the UK. Interviews with representatives with an urban interest raised

problems of crime and social disorder, physical environment, stigma and social capital. These

are discussed in more detail in the next two sections.

Regional dimensions

There are important regional contrasts within Scotland which may also shape the nature of

deprivation in different locations. The most notable of these is the east-west contrast within

the Central Belt which has grown so significantly over the last three decades (Bailey et al,

1999). Notwithstanding Glasgow’s recent economic successes, the city continues to dominate

patterns of deprivation in Scotland. Deprivation in the conurbation is characterised by high 20


levels of concentration associated with low-demand public housing estates. As a result of the

scale and nature of Glasgow’s problems, it dominates the composition of national indices.

While this gives a valid picture of deprivation nationally, it may reduce the value of national

indices for guiding policy in some other parts of the country. Within Edinburgh and the

Lothians and within Fife, for example, separate indices have been developed to identify local

problems and priority areas (ECCP, 1996; Fife, forthcoming). This approach might become

more widespread as individual Community Planning Partnerships take on greater

responsibility within their areas for “narrowing the gap” between deprived and other areas

(Scottish Executive, 2002a). Our focus remains on the construction of national measures.

3.2.2 Cultural context Deprivation, like poverty, is a relative concept, defined in relation to social norms. One

question therefore is whether Scottish people would draw the deprivation line differently to

those in other parts of Britain. There is reasonably firm evidence that there would be very few

differences. If anything, Scots would set the line marginally lower than the English, and

would be slightly more likely to emphasise material goods as necessities rather than social

activities (Pantazis et al., 2000).

Cultural factors may be implicated in behaviour associated with the risk of deprivation and

hence with the nature of deprivation. For example Scotland’s strikingly poor health record

could be partially be explained in the way that health-related behaviours (such as smoking,

drinking alcohol or poor diet) exacerbate the impacts of material deprivation (Shaw et al.,

1999). There is little substantive evidence available to support this type of argument. The

consultation process identified a number of areas where, anecdotally, a particularly Scottish

dimension may be present although these would require verification through further research.

These include: Scottish attitudes towards community and collective values, which reflect a

particularly strong focus on inclusion; a different conception of entrepreneurship and

enterprise activities; and the stigmatisation of debt and its impact on accessing credit. The

role of religion in contributing to social cohesion or division, particularly in the west of

Scotland, may be a further cultural component (Devine, 2000).

21


3.2.3 Institutional context Despite similarities in the policy frameworks of the constituent parts of the UK, there is some

distinctiveness about Scottish policy related to deprivation. In Scotland the overarching

strategy is framed in terms of social justice (Scottish Executive, 1999). The Executive has

defined social justice as “the equal and fair distribution of social values such as freedom,

income, wealth and opportunity to take part in society” (Scottish Executive, 2002a). This

suggests that social justice policy is broader than a concern with living standards but also

emphasises the centrality of poverty and deprivation for social injustice. This is reflected in

Scotland’s area regeneration initiatives, including Social Inclusion Partnerships and Initiative

at the Edge. The Executive has also developed a range of specific initiatives in other policy

areas including health (national health demonstration projects, Partners in Change), education

(community schools, the abolition of tuition fees), housing (New Housing Partnerships and

community ownership) and social work services (free personal care for pensioners). In

addition, some UK-wide policies, such as the New Deal, the New Opportunities Fund and

Sure Start, are applied to Scotland in distinctive ways. Overall, however, the picture is largely

one of similarity with other parts of the UK rather than difference.

3.2.4 Conclusions Overall this discussion suggests that, for the purposes of measuring deprivation, differences

between Scotland and the rest of the UK are limited. Approaches to the measurement of

deprivation adopted elsewhere are likely to be useful in the Scottish context. The most

significant area of difference is in the extent to which Scotland has a larger population living

in very remote areas. Differences in physical geography, culture or institutional contexts

appear relatively minor.

Views are invited on this analysis of the Scottish context and, in particular, in the

discussion of urban/rural differences.

22


3.3 Individual and area approaches

There are two quite distinct approaches to measuring deprivation – one focussed on

individuals and the other on small areas, typically wards or postcode sectors but sometimes

much smaller units such as Census output areas.

3.3.1 Individual approach The individual approach provides a direct measures of the number of people in multiple

deprivation in a given area (usually a country or region). It can also be used to examine the

characteristics of people most likely to suffer deprivation and how deprivation varies for

different groups. Recent examples would include the Poverty and Social Exclusion Survey

1999 (PSE1999) and the earlier Breadline Britain reports noted above. This approach relies

on the collection of a large amount of information about each individual, covering their

material and social circumstances.

In the past, this has only been possible through large-scale surveys. As the survey needs to

capture a wide range of information about each person, it is relatively lengthy and this limits

the number of interviews. The PSE1999 had a sample of 1530 people, enabling it to provide

robust estimates of levels of deprivation for Britain and some information on the picture for

regions such as Scotland. With sufficient resources, it would be possible to expand the survey

to provide data for sub-regions of Scotland, even local authorities, although the costs would

be considerable. Costs could be reduced by building this survey into an existing exercise,

such as the SHS, although that might limit the amount of information which could be

captured on each individual. The SHS is designed to deliver robust estimates of population

characteristics at local authority level over a two year period by interviewing over 30,000

people in that time. Data for smaller areas would be extremely costly to provide in this way.

It has been suggested that, in future, it may be possible to make greater use of the wide range

of data held by Government on individuals, their characteristics and living standards. This

includes:

• Inland Revenue data on incomes, tax and (increasingly) tax credits;

• Department of Work and Pensions data on means-tested and non-means-tested welfare

benefits;

23

• local authority data on housing and council tax benefit uptake, or on local house

conditions;


• health authority data on incidence of health problems or use of services;

• education authority data on pupils’ or students’ educational attainment; and

• police authority data on locations of crimes or victims.

Given efforts to ensure more standardised data capture procedures, it might be possible to

link records from these diverse sources to get a broad picture of each individual’s living

standards. Coverage of some aspects of deprivation is likely to be limited (social relations, in

particular) but this could be compensated by the volume of data and hence the possibility of

disaggregating results for smaller areas. There are major technical and legal challenges, not

least due to data protection legislation and individual rights to confidentiality, before such an

approach becomes a reality.

3.3.2 Area-based approach The area-based approach is designed to provide evidence of patterns of deprivation for much

smaller areas. Recent indices in Scotland have been based on wards or postcode sectors

(average population around 4000 to 5000) (Nolan et al, 2003; Gibb et al, 1998) but earlier

indices based on Census data alone focussed on output areas (average population around 100

to 150 people) (Duguid, 1995). This approach can also be used to provide evidence of the

characteristics of areas where concentrations of deprivation are greatest and of how

concentrations vary between different types of area.

To achieve this level of spatial detail without the cost of very large surveys, certain

compromises are made. First, indices are constructed from a series of indicators of area

characteristics. Some may be direct measures of a single deprivation, expressed as a

proportion of the population (unemployed, on low income or lacking qualifications) or as a

score for the area as a whole (area mortality rates). In other cases, indicators are indirect

measures, based on factors known to be associated with deprivation but arguably not

deprivations in themselves. It is important to realise, however, that neither the individual

indicators nor the index as a whole identifies the number of individuals with multiple

deprivation. In addition, while some indicators refer to the whole population, many focus on

different sub-groups (low birth-weight babies, exam passes for secondary school children, or

lack of qualifications for the working age, for example). As a result, the combined index does

not have a particular scale. It is a purely relative measure which ranks areas in relation to

each other at a given point in time.

24


One additional problem with area-based indicators concerns the definition of the areas to be

used. These are not given but have to be selected or defined. There are questions about how

big or small an area we use and about exactly where we draw the boundaries. Both decisions

can influence the picture of deprivation which emerges to some extent. This is discussed in

detail in 4.3.1 below.

3.3.3 Conclusions In the brief for this work, we were asked to consider whether one approach to measuring

deprivation (individual or area) might be preferable to the other. In our view, both approaches

may be useful as they provide different measures for different purposes. Interviewees tended

to express a similar view. Our initial recommendation would therefore be to develop both.

Given sufficient resources, it is possible that the individual approach could be used to identify

small areas with the highest concentrations of deprivation, particularly if existing government

data could be harnessed for the task. In practice, this seems unlikely at present. An alternative

might be to estimate concentrations of deprivation by combining knowledge of the

characteristics of deprived individuals from survey-based measures with data on small area

characteristics. A similar approach has been used by Bramley and Lancaster (1998) to

estimate income distributions for small areas. There would of course be significant questions

of validity with such measures but these might be at least partially addressed by a one-off

survey of deprivation in a small number of small areas.

Views are invited on this discussion of individual and area-based approaches to

measuring deprivation and, in particular, on the value of developing both types of

measure in parallel.

25


4. Developing area-based measures of deprivation

In recent years, there have been significant improvements in the availability of data for small

areas and in the methodologies used to calculate area deprivation indices. The recently-

published SIMD2003 takes advantage of both and represents a real advance. The purpose of

this section is to consider how this index might be developed in the longer term. It seeks

feedback on the indicators currently included in the index and on possible additions, deletions

or amendments. It is relatively detailed as a result and specific questions are posed on which

we seek feedback. These are not intended to limit comments. This section also covers issues

about the most appropriate units to use as well as proposing an alternative basis on which

results might be presented.

No area-based deprivation measure can hope to be comprehensive in its coverage. It will

always be possible to suggest additional indicators for aspects of deprivation not covered.

These are likely to produce diminishing returns for the cost and effort involved. Increasing

the number of indicators and domains also increases complexity and reduces transparency.

The main aim is to ensure that the measure gives broad and balanced coverage to the major

aspects of deprivation affecting significant number of people. Suggested changes will be

considered in this light.

Earlier work on area-based indices has been criticised for its reliance on indirect measures of

deprivation or “risk factors” (living in local authority housing, for example) rather than direct

measures of outcomes (living in poor quality housing). The SIMD2003 has attempted to use

develop direct measures where possible (Noble et al, 2003). We would seek to retain this

approach but we would also note that the distinction between direct and indirect indicators is

not always a simple one. The current index contains some indicators which could be

considered indirect measures.

26


4.1 Existing domains

The SIMD2003 is based on five domains, constructed from a total of 30 indicators. These

are:

• income deprivation;

• employment deprivation;

• education, skills and training deprivation;

• health deprivation and disability; and

• geographical access to services.

For each ward, domain scores are constructed from weighted or unweighted indicator scores.

The domain scores are then weighted to give an overall deprivation score. Full details of the

methods used to construct the indicators, derive weights, and produce domain scores and the

final index are set out in Noble et al (2003). These technical issues are not part of our remit.

This section reviews briefly the construction of each domain, and seeks feedback on the

current indicators.

The general questions on which we seek feedback are as follows:

• Are the individual indicators valid and sufficiently reliable?

• Are better sources of data available to update or replace any current indicators?

• Do the indicators collectively provide broad and balanced coverage of each domain?

• Do the indicators provide balanced coverage across different areas or social groups?

4.1.1 Income deprivation In the past, this domain was one of the weakest elements in area deprivation indices due to

the lack of data on incomes for small areas. The Census has never included a question on

income so levels of car ownership were used as the best available proxy. This approach has

been widely criticised for its bias against rural areas. This criticism is no longer relevant as

the SIMD2003 takes advantage of the growing availability of benefits data to provide a direct

measure of low income.

27


Two low-income groups are identified using data on means-tested benefits:

• ‘out-of-work’ – those on Income-Based Job Seekers Allowance or on Income Support

(JSA(IB) or IS); and

• ‘in-work’ – those on Working Families Tax Credit or Disability Tax Credit (WFTC or

DTC) and below 60% median income before housing costs.

The domain score is a straightforward percentage of the population on one of the four

benefits, providing an absolute measure of the numbers in income deprivation. There is only

minimal overlap between these groups according to Noble et al (2003) and this can therefore

be ignored. As separate counts are available for each ward of adults and dependent children

on each type of benefit, it would be possible to use the income indicator to provide a separate

measure of “child poverty” (the proportion of all children in a household receiving one of

these benefits).

In conceptual terms, this domain is a straightforward attempt to capture current income, the

largest single influence on poverty and deprivation. Following the discussion above, we

would be interested to explore broader approaches to measuring financial resources but we

discuss these in 4.2.1 below as a possible additional domain.

In practical terms, the main question is whether to persist in using benefits data as the best

guide to income deprivation rather than direct measures of poverty (the number of people

below a given income threshold). At present, there are no established measures of income

distributions for small areas. Although some work has been carried out (Bramley and

Lancaster, 1998), developing robust and widely accepted measures would be costly. Gaining

access to data held by the Inland Revenue would help although this is likely to be difficult.

By contrast, data on benefits is already available for small areas on a consistent basis. This

seems the strongest route to pursue but we would welcome comments on this.

28

Five more specific issues arise from the use of benefits data. First, Noble et al (2003) note

that the reliance on claimed benefits rather than eligibility is problematic as differences in

benefit take-up rates may distort the picture. This has certainly been a concern from the rural

perspective (Countryside Agency, 2003), reflecting a belief that people in rural areas are

more likely to under claim, due to lack of awareness of entitlements or reluctance to claim

(Shucksmith et al, 1994, 1996). Using data from 1996, Bramley et al (2000) show that there

do indeed appear to be differences in benefit take-up rates across Scotland, although these are


much greater for older people (above pension age) than others. For older people, take-up is

significantly lower in more affluent areas. Additionally, take-up is slightly lower in more

rural locations but this difference largely disappears once levels of affluence are taken into

account. In other words, the problem seems to be largely associated with the affluence of

rural areas (on average) rather than their rurality per se. Bramley et al (2000) provide

estimates of local eligibility rates by adjusting benefits data for non-take-up. This reduces

absolute differences between areas in terms of income deprivation but it has very little impact

on the overall picture of the distribution of poverty across Scotland. For individual areas, of

course, the changes might be more significant. Our view is that it would be worthwhile

updating this exercise to develop local eligibility indicators for use in future indices, not least

because these would provide a more accurate measure of the absolute scale of income

deprivation. A wider range of benefits could also be covered.

A second issue is whether the available benefits data enables us to capture problems of low

income across the working age population. The inclusion of both out-of-work and in-work

low income groups is a positive feature of the current index. It means that the measure

captures some of the problems which may arise from underemployment or from employment

in low wage industries, often said to be a particular problem in rural areas. The reliance on

WFTC and DTC, however, means that coverage of low income in-work groups is restricted

to those with children and/or with disabilities. From April 2003, however, WFTC has been

replaced by two new tax credits – Working Tax Credit and Child Tax Credit (WTC and

CTC). The former is available to all low income households, regardless of whether they have

children or not. As such, the move to use WTC would broaden coverage of low income

groups of working age and strengthen the index.

Third, current measures miss older people (above retirement age) on low incomes but not in

receipt of Income Support (IS). Again, there would be a concern, especially from a rural

perspective, that this might lead to an undercounting of income deprivation in some locations.

With the introduction of the Minimum Income Guarantee and soon the Pension Credit (from

October 2003), there should be better data on low income pensioner households above IS

levels. The Pension Credit has potentially very wide coverage (up to 57 per cent of

pensioners in Scotland) so a low income threshold may need to be applied.

29


With improved coverage of working age and older people from these new benefits, there is a

case for producing separate measures of income deprivation for child, working age and

pensioner alongside the overall domain score.

Data on housing or council tax benefits might give a broader picture of low income for both

working age and older households, provided duplicate cases could be eliminated. This data

was explored by Noble et al (2003) but was not felt to be of sufficient quality to include in

the domain at present.

Fourth, a different aspect of income deprivation not captured in the current index is duration

on low income. A number of studies have argued that people who spend a prolonged period

on low income are likely to face more severe problems of deprivation (Nolan and Whelan,

1996). One approach to measuring this would be to use benefits data to identify those who

had been on low incomes for prolonged periods (say, more than one year). In calculating the

income deprivation for each ward, this group could be given an additional weighting to

reflect the severity of their problems. The drawback of this approach is that the income

indicator would no longer represent simply the percentage of people in each area in income

deprivation although it might be possible to publish that figure separately.

A fifth issue is the local cost of living. Obviously the same income may produce different

standards of living depending on the cost of housing, basic goods or services and so on. One

view is that the cost of many basic goods and services, together with transport costs, tend to

make living in remoter areas more expensive (Shucksmith et al, 1996) but others might see

urban living costs as higher, particularly given housing costs. It would certainly be useful to

examine this issue in more detail to see how significant a problem it is. If data were available

on local cost of living and on local income distributions, the problem could be addressed by

applying a local income deflator to obtain more accurate measures of income deprivation.

With the current measure based on the proportion of the population in receipt of specific

benefits, adjustments would need to be made in a different way.

Views are welcomed on the income deprivation indicators, particularly on the issue of

direct estimates of poverty versus benefits data but also on the more detailed issues

identified.

30


4.1.2 Employment deprivation According to Noble et al (2003: p.14), this domain aims to measure “enforced exclusion from

the world of work”. This is distinct from the income deprivation which such exclusion may

lead to. There are four indicators in this domain. Two relate to unemployment and two to

sickness and disability:

• unemployment claimant count of those aged under 60 (Job Seekers Allowance);

• compulsory New Deal participants, not included in unemployment claimant count;

• Severe Disablement Allowance recipients aged under 60; and

• Incapacity Benefit recipients aged under 60.

As with income, the domain score is the percentage of adults receiving these benefits and is

then a direct measure of the numbers in employment deprivation.

The conceptual basis of this domain is a little unclear. It is set out as a collection of direct

indicators of “employment deprivation”. People experiencing employment deprivation are

defined as “those who wish to work but are prevented from doing so through unemployment,

sickness or disability” (Noble et al, 2003: p.14). A minor point is that “unemployment” is the

result of being unable to find work, not the cause. The type of factors which might make it

more difficult for someone to find work include personal factors such as human capital

(skills, work experience), demographics (age, gender), household status (caring

responsibilities) but also employment demand in the area and mobility. Human capital factors

fit better in the education, skills and training domain and we discuss employment demand and

mobility (access to employment) as a possible additional domain in Section 4.2.5 below.

More importantly, it is unclear that not having employment is a deprivation in the sense

discussed in Section 2. This approach seems to value employment for its own sake rather than

as a means to other ends (primarily income, but also status and social contact). If it is the

latter which are more important, then these should be captured directly. The current approach

would fit well with some conceptions of social exclusion which emphasise access to

employment (for those who want it) as necessary for inclusion (see 2.4). Comments on the

approach taken in this domain would be welcomed.

Taking the domain as it stands, the indicators take a much broader definition of

unemployment than either the claimant count or the International Labour Office (ILO)

measure, captured by the Labour Force Survey. This is to be welcomed, given the work on

“hidden unemployment” (Beatty et al, 2002). It is worth noting, however, that there are

31


strong correlations between different measures of unemployment (Webster, 2000). This

implies that the use of one rather than another would have little impact on the relative

position of different areas (although the absolute positions would change).

One issue discussed by Noble et al (2003) is the treatment of unemployed lone parents.

Should they be regarded as involuntarily unemployed and therefore included in the count of

“employment deprived” or as voluntarily unemployed (inactive), and therefore excluded?

Noble et al also recognise that there are others who may be seeking work but who are not

registered unemployed or long term sick. They discuss the partners (usually female) of JSA

claimants who are not required to sign-on themselves but this group might also include

people (usually female) whose partners are in employment but who are not working because

of caring responsibilities. Some of these make a positive choice not to work but, for others,

the choice is at least partly conditioned by judgements about the type of work they might

hope to obtain. Again, the impact on the relative position of each area is likely to be small but

the current approach leads to an undercount of people, particularly women, “excluded from

the labour market”. We would welcome comments on this issue.

One criticism of all the employment indicators is their reliance on figures for a single month.

For some areas, notably rural areas with fluctuations due to seasonal employment

(particularly in tourism and agriculture), this can give a potentially misleading impression. It

would be relatively simple to extract data for one month in each quarter to give an average

figure for the year without seasonal bias. We would recommend that this is done in future.

A different criticism is that the indicators do not capture aspects of employment deprivation

other than “unemployment”. These are no measures of underemployment (part-time or

seasonal working) or of low quality of employment (job insecurity, poor working conditions,

etc.). Suggestions on direct measures of these or related matters would be welcome although

these would need to separate those with poor conditions resulting from a weak labour market

position from those whose poor conditions of work are compensated (e.g. oil-rig workers). It

should be noted that low pay issues are picked up through Income Deprivation indicators for

low income in-work groups.

Views are welcomed on the current employment deprivation indicators, both on the

conceptual basis for this domain and our specific comments on current indicators.

32


4.1.3 Health deprivation and disability This domain seeks to identify areas with relatively high proportions of people who die

prematurely or whose quality of life is impaired by poor health or disability (Noble et al,

2003: p.16). The indicators (with weightings in brackets) are:

• Comparative Mortality Factor (CMFs) for under 75s (13 per cent);

• hospital episodes related to alcohol use (12 per cent);

• hospital episodes related to drug use (11 per cent);

• Comparative Illness Factor (CIF) (39 per cent);

• emergency admissions to hospital (10 per cent);

• proportion of population being prescribed drugs for anxiety or depression or

psychosis (8 per cent); and

• proportion of live singleton births of low birth weight (<2,500g) (7 per cent).

This domain incorporates a number of improvements in measuring health deprivation and

disability over previous indices, notably: the move to broader measures of chronic ill health

(rather than narrower measures of the incidence of certain acute conditions); the use of

hospital episodes data to capture health effects of alcohol and drug misuse (rather than the

use of alcohol and drugs per se); and efforts to capture problems of mental health.

In conceptual terms, poor health can be seen as a deprivation in its own right and also a

contributory factor to other forms of deprivation (unemployment or low income, for

example). This is not to argue that everyone with an illness is deprived but to recognise that

one of the major determinants of health inequality is material inequality (Shaw et al, 1999).

This implies that the aim of this domain is to identify the excess ill health or mortality

attributable to deprivation. The health indicators should therefore be seen as indirect

measures. Simple measures of ill health or death would identify areas with high numbers of

older people so there is a need to control for known demographic effects to bring out the

deprivation effects. Two indicators (Comparative Mortality and Morbidity Factors) adopt this

approach, producing rates which are standardised for age and gender. In contrast, the other

five health indicators have not been standardised. It would be useful to explore whether

population structure has a significant effect on these and to consider controlling for this if

necessary.

33


The Comparative Mortality Factor (CMF) has been limited to people under 75. The argument

for doing this is not explicit but presumably relates to the fact that social effects are greater

for younger age groups than older. Nevertheless, social effects persist through older groups

and there seems little reason for limiting the indicator therefore. This age cut-off also leads to

a possible bias against women given their greater longevity. One proposal would be to

construct two mortality indicators to improve population coverage and to provide the

possibility of disaggregating into working age and post-retirement groups.

Previous efforts to measure chronic ill health have used the 1991 Census question on limiting

long-term illness (LLTI) as this provides a direct measure for small areas. The Comparative

Illness Factor was constructed from benefits data because the Census was out-of-date and, in

any case, could only be updated every ten years. It is a less comprehensive measure, being

based on estimates of the number unable to work due to ill-health (IB or SDA claimants) or

in receipt of additional support to meet care or mobility needs (DLA or AA). With

publication of the results of the Census 2001, it would also be useful to explore the extent to

which the two approaches produced the same results, as a means of validating the current

indicator. An alternative source for a broader measure of morbidity might be the Continuous

Morbidity Recording system which has been developed in Scotland. This draws on records of

visits to GPs, distinguishing first and subsequent contacts for a given problem. Data is

available on patients’ home address so it would be possible to construct morbidity rates from

this. One limitation might be the present coverage of the system (around 7 per cent of GPs).

An indicator on smoking was considered for the SIMD2003 but rejected due to lack of

adequate data. This might be worth exploring further. Smoking is a considerable public

health problem because it is seen as the most preventable cause of ill health in the country.

Within Scotland, smoking rates are several times higher among people in unskilled

occupations than among those in professional occupations (Scottish Executive, 2000a). Data

on women smoking at the outset of pregnancy is available for small areas. While it may offer

a good proxy for smoking behaviour more generally, recording errors meant it was not

included in the SIMD2003. Comments on alternative sources of data on smoking for small

areas would be welcomed.

Views are welcomed on the conceptual basis of the health domain and on the indicators

used to construct it at present.

34


4.1.4 Education, skills and training deprivation The domain incorporates measures which apply to secondary school age, post-compulsory

school age and working age groups. As such, it takes a broader approach to education, skills

and training than many previous indices which is to be welcomed. The current set of

indicators (with weightings) is as follows:

• working age adults (25-59) with no qualifications (13 per cent);

• post-compulsory school age not in full-time education (made up of two indicators

which share the same denominator: pupils age 16-18 not in full-time education; and

proportion of 17-19 population who have not successfully applied to Higher

Education) (16 per cent);

• pupil performance on SQA at Stage 4 (65 per cent); and

• secondary level absences (6 per cent).

Conceptually, the focus of this domain is ambiguous. On the one hand, it is seen as

measuring “the key educational characteristics of the local area that might contribute to the

overall level of deprivation and disadvantage” (Noble et al, 2003: p. 20 – emphasis in

original). In this perspective, the education indicators are risk factors increasing the

probability of an individual suffering deprivation. Hence they act as indirect measures of

deprivation. On the other hand, low educational achievement is seen as a deprivation in its

own right: “if - for whatever reason - one area has better educational results than another that

may be less disadvantaged in other respects, then this area is less educationally deprived”

(p20 – emphasis in original). This sees education as an end in itself and as contributing to

someone’s quality of life. In this sense, the indicators may be seen as direct measures of

deprivation.

Looking at the indicators, the focus on formal qualifications and on working age groups

suggests that the main focus is employability. This perception is reinforced by the very high

weighting given to pupil performance. The justification for including absence rates is not

given but could also be linked to their relationship with exam performance and hence

employability. Absence rates may also be seen as an indirect indicator of deprivation, given

what is known about the relationship between absence and family circumstances (SEU,

1998).

35


As the report on the SIMD2003 notes, there is some debate over how educational

performance should be measured but a focus on formal qualifications has the advantage that

these are not only valued educationally but also in the job market. Data is also readily

available. We support this approach but would welcome comments on this as well as

suggestions of alternative measures of educational outcomes or deprivations.

Four specific issues arise in relation to the current indicators. First, in relation to pupil

performance on SQA at Stage 4, one recommendation has been to use an average over three

years, rather than relying on one cohort. This seems straightforward provided the

denominator is treated in the same way.

Second, the data on absence rates are rather weak. The indicator is derived by averaging

absence rates for secondary schools back to the home locations of pupils, as assessed through

home addresses of those taking SQA4. At present, it contributes just 6 per cent of weight of

the final domain score. We would welcome comments on the value of retaining this variable

and on improving its construction.

Third, the indicator for working-age adults lacking qualifications is based on rather weak

data. The current method models local authority estimates from LFS data to ward level using

the Census 1991. Updating the figures using Census 2001 data would clearly help and this

could be done quite easily now that data has been published. This method could also be

checked against direct ward-level measures derived from Census 2001 (although the latter

cannot be updated in future). Suggestions for better sources of data on qualifications for

working-age adults would be welcomed.

Fourth, from a policy perspective, there has been increasing attention paid to more the basic

skills of literacy and numeracy (Scottish Executive, 2001a). For measuring deprivation, these

would have the value of being more clearly related to individual quality of life as well as

employability. We would welcome comments on the value of adding indicators and on

possible sources of data for small areas.

Views are welcomed on the conceptual basis of this domain and on the indicators used

to construct it.

36


4.1.5 Geographical access to services This domain is designed to capture the effects of living in locations which lack key services

locally – usually, but not always, smaller settlements or rural areas. Poor access may have a

number of impacts including higher costs in travelling to use services or less frequent usage,

and both may be considered as direct deprivations. The six variables cover access to health

care, food, education, fuel, financial services and communications. The indicators in the

current domain (with weightings in brackets) are:

• road distance to a GP/health centre (19 per cent)

• road distance to a general store/supermarket (25 per cent)

• road distance to a primary school (12 per cent)

• road distance to a petrol station (11 per cent)

• road distance to a bank/building society (18 per cent)

• road distance to a community internet facility (15 per cent).

This domain has drawn strong criticism in the past (Robson et al, 2002) because it has a

significant negative correlation with all the other domains and with the overall deprivation

index. This is not necessarily a problem. After all, the aim of the geographical access domain

is to highlight an aspect of deprivation not previously captured by the index. More generally,

if one starts from the view that deprivation is multidimensional, then one would not

necessarily expect different indicators to be highly correlated. Even so, the strength of the

negative relationship is a cause for concern.

A more conceptual criticism is that poor geographical access is not in itself a deprivation but

is a risk factor for deprivation. Not everyone living a in low access area is deprived. This can

be linked to another criticism, that poor geographical access may be associated with more

positive outcomes. Some people choose to move from higher access urban areas to lower

access suburban or rural areas in search of a better quality of living environment - both

physical and social. Measuring geographical access without measuring quality of living

environment would therefore lead to biases in the index. Arguably, the problems of poor

environment are most strongly associated with large cities and former industrial towns of the

Central Belt, hence the pressure for out-migration to suburban and “accessible rural” areas

from these locations. It is less clear that there is a continued improvement in quality of

environment as one moves from the latter to increasingly remote rural locations. As such, the

37


biases of measuring geographical access along may not be so great. We would welcome

comment on the conceptual basis of this indicator.

The other major criticism of geographical access is that it provides only a very partial

measure of the extent to which people are able to use particular services or of the quantity or

quality of service available. As Robson et al (2002) argue, it may well be preferable to be five

miles from a good school than one mile from a school regarded as failing. Midgley (2003)

points out that geographical access may be the least of the factors which determine the ability

of an individual to secure and use financial services, whether they live in rural or urban

locations. Bramley and Ford (2001) found that deprived households in Britain (as defined in

the PSE1999) were more likely to face poorer services or be constrained in their use of

services due to cost or availability. Measuring these wider aspects of access to services would

be challenging, particularly for small areas, but we would welcome comments on whether the

focus of this domain should be broadened and how this might be achieved.

Turning to the current set of indicators, one issue is whether they give an appropriately

balanced measure of geographical access. Four of the current set (doctors, supermarkets,

banks/building societies and petrol stations) are justified in the SIMD2003 report because the

vast majority of households use these services (Noble et al, 2003, quoting results from

Bramley and Ford, 2001). As daily attendance at primary schools is compulsory, the same

argument would applies there. Access to petrol stations might be considered inappropriate,

given that low income households are much less likely to have access to a car. Bramley and

Ford (2001: p11) show that the highest income group are 76 per cent more likely to use a

petrol station than the lowest income group. Access to community internet facilities also

seems a poor indicator of communications given that relatively few people are likely to use

such facilities. Access to a post office has been omitted although it has been used elsewhere

and, according to Bramley and Ford (2001) is also used by the vast majority and seen as an

essential local service. It is also likely to be a service of particular importance to many on low

incomes. We would welcome comments on the range of services included in this domain.

A second issue concerns the measurement of accessibility. The SIMD2003 improves on

previous measures by using road distance between two points rather than direct (“crow flies”)

distance, but the measure still does not accurately identify how difficult it is to travel between

two points. One suggested improvement would be to use drive times, not least because that

approach was used to construct the urban-rural classification used in the SHS. Another 38


improvement, noted by Noble et al (2003), would be to take account of the availability of

public or private transport. Alternatively “generalised travel costs” could be used. These

reflect physical distance but also: type of transport available; time taken; and convenience or

quality of transport. There are also specific problems associated with measuring distance to

services for those living on islands lacking a particular service (Noble et al, 2003). We would

welcome comments on these issues.

Views are welcomed on the conceptual basis of this domain and on its construction at

present. Views on extending the domain to capture broader aspects of access to services

would be particularly welcomed.

39


4.2 Other domains

This section discusses potential domains that might be introduced to the area-deprivation

measure in future. Bearing in mind the comments at the start of this section, our aim is to

identify significant aspects of deprivation not yet captured by the index. The discussion

should not be limited by the availability of data but should seek to identify the most important

gaps.

The first proposed addition is on financial resources, in an attempt to go beyond narrow

measures of current income. This also covers issues of financial exclusion. The next three

were all considered for inclusion in the SIMD2003 but were not included in the final index

due to lack of suitable data. These are all direct aspects of deprivation, measuring elements of

individual living standards which reflect levels of income: housing, the physical environment,

and crime and social order. In addition, we discuss the value of adding domains to cover

geographical access to employment and communications. The latter includes issues of

“digital exclusion”. Finally, we consider a domain to capture social aspects of deprivation,

covering social relations, social capital and social status. We outline the scope of each

potential domain, its conceptual basis and the measures which might be adopted.

We would welcome feedback based on the following:

• Are these additional domains sufficiently important to warrant inclusion?

• Which domains should receive priority?

• How should they be conceptualised? What are we trying to measure and why?

• What indicators should be used to construct balanced measures for these domains?

• What data sources might be investigated?

40


4.2.1 Financial resources and financial exclusion As noted in Section 2.2 above, current income may give only a partial measure of the

financial resources available to an individual. We are therefore proposing an additional

domain to capture other aspects of financial resources more broadly. We are basing this

argument on Nolan and Whelan’s work in Ireland. Before proceeding, it would be important

to carry out a similar exercise in Scotland to assess the extent to which the relationships are

the same here.

First, this domain could cover levels of savings and debts. There are important definitional

issues to consider, such as the treatment of certain types of debt (mortgages, for example) or

savings (housing equity or pension funds). Second, this might measure access to and take up

of financial services such as banking, credit cards or insurance. Lack of such services

(financial exclusion) can lead to a higher cost of living, lack of access to other services or

goods and higher financial risks for households (HM Treasury, 1999; Goodwin et al, 1999).

Third, a measures could capture the extent to which individuals have experienced financial

stress more directly. This would include measures of the number of households who have

experienced problems in paying for basic items (utility bills or housing costs) or who have

reduced their consumption of basic items for financial reasons.

There are no sources of data for small areas on any of these matters. Some survey data exist,

particularly through the SHS and the PSE1999, although the reliability of such information

would need to be verified before it was used to model small area data.

Conclusion

We believe this is potentially an important domain, where further work would be justified.

Our intention would be to try to build a broad picture of access to financial resources but we

would be keen to build this on a better understanding of the relationships between income,

other financial resources and deprivation in Scotland.

Views on the value of this domain would be welcomed, along with comments on how

data might be obtained for small areas.

41


4.2.2 Housing deprivation The inability of afford adequate housing may be a major source of deprivation. Poor housing

may also impact on a range of other dimensions, such as health. The SIMD2003 does not

include a housing deprivation domain at present due to the lack of adequate data. This is an

important gap and one which it should be possible to begin to fill quickly. In the PSE1999,

housing issues were amongst the items most likely to be considered necessities of life by

people in Britain. The top three items (with the proportion of people considering them

necessities in brackets) were:

• beds and bedding for everyone (95 per cent);

• heating to warm living areas (94 per cent); and

• damp free home (93 per cent) (Pantazis et al, 2000).

The starting point needs to be a discussion of what constitutes housing deprivation. Different

approaches have been adopted in England and Wales. In the Welsh index of deprivation, the

housing domain was narrowly defined in terms of “housing stress”, measured in terms of

physical conditions: proportion of housing in disrepair; proportion of houses without central

heating; and proportion of houses lacking roof/loft insulation (Noble, Smith, Wright et al,

2000). For the English index, a broader approach was adopted, covering physical conditions

and the suitability of housing for occupants. The indicators were: poor quality private sector

housing; household overcrowding; and homeless households in temporary accommodation

(Noble, Smith, Penhale et al, 2000).

In their initial report on updating the English index, Noble et al (2002) question whether

physical conditions and suitability should be seen as two separate issues and covered by

different domains. Our view is that both types of problem can be seen as stemming from the

same basic cause (lack of resources to secure adequate housing for an individual’s needs) and

that it would be appropriate to keep these within one domain.

Physical conditions

42

The obvious starting point in defining minimum physical conditions are Government

standards. In Scotland, the current minimum standards are defined by the Tolerable Standard

and the list of basic amenities. This has limited value as a guide to housing deprivation

because it represents a minimum “condemnatory” standard rather than measuring

habitability. A new standard is being developed (Housing Improvement Task Force, 2003).

The current proposals cover the following:


• compliance with the Tolerable Standard;

• free from serious disrepair;

• energy efficient;

• equipped with modern facilities and services;

• healthy, safe and secure for their occupants;

• adapted, or capable of adaptation to meet particular and/or special needs.

Some aspects of physical conditions are available from the Census: the presence of standard

amenities and central heating. Other factors, such as thermal efficiency, may require costly

surveys. The Scottish House Condition Source would be a starting point but was not designed

to provide estimates even down to local authority level.

Care would need to be taken that the indicators of poor physical conditions did genuinely

capture the situation of deprived households forced to live in unsatisfactory housing due to

lack of income rather than the choice by some more affluent households to buy older

properties in poor condition or lacking modern amenities.

Suitability

Suitability for individual needs can cover a wide range of issues. Data on overcrowding could

be obtained from the Census 2001 although this would be difficult to update. A second aspect

would be homelessness. This would cover homelessness in many forms, from rough-sleeping,

through living in short-stay hostel or temporary accommodation, to living with friends or

family on a temporary basis. The scale of these phenomena is often difficult to measure.

Measures for small areas may reflect levels of provision (the supply of hostels or temporary

accommodation) rather than the places where homelessness originates. Administrative data

on homelessness may be biased by local administrative procedures. As Noble et al (2003)

noted, local authorities are now obliged to collect information from homeless applicants on

their last fixed address. This could provide some useful information on origins, although it

would only cover those who made contact with the local authority. Further issues would

include the extent to which housing had been adapted to meet particular needs of the

occupant, for example, in relation to impaired mobility and the need for ‘barrier free’ design.

43


Costs, affordability and access

The cost of housing can impact on deprivation in a number of respects. First, as housing is a

major item of expenditure for most households, it can be a key determinant of variations in

the cost of living, as discussed in 4.1.1. Second, high costs can lead households to reduce

their consumption of housing, contributing to overcrowding. Interestingly, the Welsh index

rejected overcrowding measures as they were potentially distorted by property prices. The

opposite argument would be to include overcrowding precisely to capture this effect. Third,

high housing costs limit access to housing in the first place, leading to homelessness. Again,

this effect can be captured to some extent.

Arguments have been made for including house price or affordability (price/earnings)

measures in deprivation indices (Countryside Agency, 2003). There are a number of

problems with this, however. First, high house prices may indicate problems of access for

low income groups but they also reflect the nature of the housing stock and the incomes of

households seeking accommodation in the area. Indeed, areas with high levels of deprivation

are more likely to suffer from very low house prices. Second, there is a question of the spatial

scale at which to measure access to housing. Is it reasonable to argue that someone is

deprived of access to housing when prices in their ward are high but there is affordable

housing in a neighbouring ward? How wide an area should we expect households to search

over? Third, high house prices appear to be more of an area-level variable rather than a direct

measure of deprivation (such as overcrowding or homelessness).

Conclusion

Access to adequate housing is a key priority for people and a major gap in the current index.

We would conceptualise housing deprivation in terms of physical conditions and in terms of

the ability of individuals to secure housing suitable for their needs. In the short-term, some

aspects of housing deprivation can be captured from Census data. New data on origins of

homeless applicants would also be worth exploring. Further developments may be possible

following release of housing data in the Neighbourhood Statistics service. We do not think it

would be useful to develop measures of cost or affordability.

Views on the conceptualisation of housing deprivation and on possible indicators would

be welcomed.

44


4.2.3 Physical environment A poor physical environment can have a direct impact on quality of life. It can be seen as a

deprivation as financial resources are a major determinant of an individual’s ability to secure

housing in areas with a reasonable quality of environment. Some aspects of the physical

environment may also contribute to poor health outcomes and to the stigmatisation of people

who live in the area. The physical environment covers issues such as air, water or noise

pollution as well as the visual quality of the environment. This domain was not included in

the SIMD2003 due to lack of adequate data (Noble et al, 2003).

To date, it has not proved possible to develop adequate measures of the quality of air and

water, or noise levels. The Scottish Environmental Incident Surveillance System (SEISS)

carries out monitoring of environmental incidents likely to have immediate health

implications (chemical or radiation leaks or water contamination, for example). The focus on

acute incidents is unlikely to produce useful data for measuring deprivation as there is no

clear link between such incidents and poverty. A more useful approach would be to measure

air, water or noise quality on an on-going basis, on the assumption that lower income groups

are less able to avoid these negative factors. The problem to date has been limitations of data.

It is difficult to see how an overall assessment of visual quality could be developed, not least

because the criteria to be applied would be so subjective. Is a view of lochs and mountains

better or worse than a view of Georgian townhouses or of factory sheds? The issue is

simplified if it is restricted to clearly negative features, such as graffiti, litter or vandalism,

although there may be some overlap with the crime and social order domain. The general

appearance of a neighbourhood can have a major impact on residents’ satisfaction with where

they live (Parkes et al, 2002). Views might be obtained from residents (through the SHS, for

example) or from “objective” assessors who make judgements on a wide range of areas.

Conclusion

There was some support for developing indicators to cover this domain, with key issues the

presence of “negative” features such as graffiti, litter or vandalised property.

Views on the conceptualisation of physical environment and on possible indicators

would be welcomed.

45


4.2.4 Crime and social order Both interviewees and focus group participants emphasised the importance of crime and

social order as factors shaping quality of life and previous research reinforces this (Parkes et

al, 2002). Living in an area with high levels of crime or social order problems has a direct

impact on quality of life. As with physical environment, lack of income is a major factor

constraining some individuals to live in areas with high crime or social disorder problems and

these can therefore be seen as problems of deprivation.

Various approaches to measuring crime and social order problems are possible: police data,

other data on the incidence of crime or victimisation, and data on individual perceptions of

local problems or the risk of crime.

Although a great deal of crime data is gathered by the police, very little is currently available

at the small area level due to recording practices and the confidentiality of some information

(Noble et al, 2003). Steps to standardise recording practices are underway and these might

make it possible to construct ward level statistics although this appears to be someway off. In

the meantime, Robson et al (2002) have proposed a detailed methodology for England to

produce local crime statistics from beat statistics and this might provide a model for Scotland.

Noble et al (2003) also report that fire service data on malicious fires or false call-outs might

provide some insight into disorder problems although they were not able to access that data in

time for the SIMD2003.

Many minor crimes and low level public disorder problems go unreported, particularly in

areas where insurance coverage is low. Self-reported victimisation data addresses some of

these problems for personal or property related crimes. At national level, data is available

from various sources including the Scottish Crime Survey and the SHS, although neither

provides details at a small area level directly. Insurance premiums have been used as proxies

for property crime levels in the past but these have been widely criticised, not least for

providing relatively insensitive indicators (Noble et al, 2000).

Data on individual perceptions of problems with local disorder and on fear of crime are

available from sources including as the SHS, although again these do not provide direct

measures for small areas. It can be argued that it is the fear of crime that actually deprives

individuals from undertaking activities and opportunities in their lives. Trends in the levels of

fear of crime do not always reflect changes in the official crime rates in local areas. 46


Conclusion

There is strong support for developing small area measures of crime and social order as these

are seen as major determinants of quality of life. Direct measures of crime levels for small

areas might be derived from police and fire service data, although there are some questions

over quality at present. Data on self-reported victimisation rates or individual perceptions of

fear of crime or of specific problems might be modelled to small area level.

Views on the conceptualisation of crime and social order and on possible indicators

would be welcomed.

4.2.5 Geographical access to employment The SIMD2003 already contains a measure of geographical access to services to reflect the

problems which people in some locations have in travelling to access basic services. Similar

problems face people in travelling to work although the details are rather different. For this

reason, we suggest a separate domain to cover access to employment although it might be

possible to include it within the existing geographical access domain. Alternatively,

following the discussion in 4.1.2, this might be an indicator to include in the employment

deprivation domain.

There is an extensive research literature on the problems of geographical access to

employment, from the US in particular (Ihlanfeldt and Sjoquist, 1998). This shows that

access to employment is determined by a number of factors. First, it is determined by where

people live. Some locations have more jobs nearby ('job rich' locations) while others have

relatively few ('job sparse' locations). Second, it is determined by the level of competition for

those jobs. Services are generally “non-rival goods”; many people can use a bank, petrol

station or shop at the same time. Jobs are “rival goods” as only one person can occupy each

position at any one time. Access to employment is therefore influenced by the level of

competition relative to the number of jobs available. Third, access to employment is

determined by individual characteristics, particularly skills or experience. It is the availability

of job opportunities suited to an individual’s skills (and of competition from similar workers)

which determines access. Fourth, access in a function of mobility. People reliant on public

47


transport tend to find that fewer employment locations can be reached for a given cost,

reducing their access to employment.

People with low access to employment will be faced with higher costs in travelling to work

on average, reducing their net income. This may result in higher unemployment rates. In

addition, they may face problems in getting information about job opportunities

(Immergluck, 1998). There is also more anecdotal evidence that employers may prefer

workers who live nearer to the workplace, believing that shorter commuting distances are

associated with greater reliability and punctuality.

A range of methods for calculating geographical access to employment have been developed,

although the most sophisticated are those of Shen (1998). These allow for the incorporation

of residential location, employment location, competing workforce, occupational differences

and differences in mobility. They are quite demanding in terms of data requirements but

could be operationalised at small area level using the Census. Some of the data could be

updated between Censuses. Workplace employment data is available for small areas from the

Annual Business Enquiry and local authority-level data from the Labour Force Survey could

be combined with small-area population figures to provide some updating of workforce

characteristics.

Conclusions

Conceptually, this could be seen as a separate domain, or as part of either the geographical

access to services domain or the employment domain. It is likely to produce rather different

results than the services domain, not least because of the need to account for competition

between workers for jobs. There is a substantial research literature on the subject and Census

data is available to produce a sophisticated measure which could be partially updated on an

annual basis.

Views on the conceptualisation of the geographical access to employment domain, on its

relationship with other domains and on the construction of the measure would be

welcomed.

48


4.2.6 Communications and digital exclusion The emphasis on geographical access gives only a partial impression of the ability of

households to obtain goods or services. There is increasing use of telephone, television and

Internet to provide information and advice, and to deliver services including entertainment,

retailing, and financial services. There is also concern over a growing “digital divide”

particularly as those least able to take advantage of the Internet tend to be those who are

already the most excluded in society (Scottish Executive, 2001b). A useful complement to the

spatial measures might be measures of access to communications technologies.

There was some support for this domain from interviewees, especially in relation to “digital

exclusion” issues (access to computers and the Internet). Wider public opinion seems to be

focussed on the more basic means of communication. In the PSE1999, 71 per cent thought a

telephone was a necessity, and 56 per cent saw a television as such. Other items got less

support: a daily newspaper (30 per cent); home computer (11 per cent); mobile phone (7 per

cent); access to the Internet (6 per cent); satellite television (5 per cent).

Measures of access to means of communications might therefore start with telephones and

televisions. Some data is available from surveys such as the SHS and could potentially be

modelled to small area level. For telephones, service-provider data would probably be

difficult to obtain due to the fragmented and competitive nature of the market. For

televisions, ownership is near universal and, for the rest, it would be essential to distinguish

those who choose not to own a television from those who could not afford one. In the future,

measures might examine access to interactive services delivered through the television but it

is difficult to see these as necessities at present.

Conclusions

This is potentially a useful addition to existing domains but it is not clear at present what the

focus should be. We would welcome suggestions which might help develop it further.

Views on the conceptualisation of the communications and digital exclusion domain and

on the construction of indicators would be welcomed.

49


4.2.7 Social relationships, social capital and social status Deprivation measures, particularly area-based measures, have tended to focus on material

aspects of the problem as these are most easily captured. This is true for the SIMD2003 as

well as its predecessors. In the framework in Section 2, we were clear that social relationships

are just as much “necessities of life” as basic material goods or services. This view was

supported by many comments in the focus groups and interviews. The PSE1999 found that

the majority of people supported the idea that the ability to visit friends and family, to have

people round for a meal, to buy occasional presents or take part in celebrations should all be

seen as necessities. Money to keep up a hobby or leisure activity was also seen as a necessity

by 78 per cent. The aim of this domain would therefore be to capture the extent to which

people suffer from social deprivation as a result of lack of financial resources.

Social deprivation may also have further consequences. At an individual level, the loss of

social ties may lead to a loss of social support, isolation and stress (Burns et al., 2001;

Wilkinson, 2000). At a collective level, the loss of social ties may lead to a loss of “social

capital” with a range of possible consequences discussed in 2.4 above. Debates about whether

social relations or social capital bring further benefits are secondary. If these benefits exist,

they should be captured by other domains.

The problem with this domain is deciding what to measure. The discussion of social capital in

Section 2.4 highlighted the fact that there are different types of social relationship which may

have different uses or values for individuals. One approach would be to look at personal

relationships. Initial analysis of SHS data (undertaken for this project) suggests that the

number of contacts an individual has with family and friends declines as income falls. This

fits with the idea that lack of income acts as a barrier to maintaining and developing social

relationships. A more common view, however, might be that people on low incomes tend to

have more and stronger contacts with close family as these may become increasingly

important as sources of practical or financial support.

A second approach would be to look at other types of relationship. Much writing on social

capital stresses wider sets of connections, including those arising from participation in

voluntary association or activities or civic engagement, such as voting. These activities might

be equally important for people living in poverty and, as noted, a majority in Britain would

agree that people should be able to participate in hobbies or leisure activities.

50


A third approach would look less directly at personal relationship and more at issues of

stigma. Concern over the ways in which residents from particular areas were stigmatised and

marginalised by virtue of the image or reputation on the area in which they live. This came

through strongly in the focus groups with community representatives. It has a direct impact

on individuals’ quality of life and may also impact on their ability to find work or to use

certain services (Dean and Hastings, 2000). Perceptions of feeling stigmatised would need to

be measured through the views of residents themselves.

Conclusion

Developing a measure of social deprivation is seen as very important but further work is

needed to identify the types of relationship which people see as important and how these are

affected by declining income. Further work would also be needed on the issue of

stigmatisation and how this is dealt with.

Views on the development of conceptual basis of this domain would be welcomed.

4.3 Units for analysis

One specific question for this project is the most appropriate units to use for measuring

concentrations of deprivation. A number of issues need to be considered. First, there is the

question of size. Across the board, there was a demand from interviewees that the index use

the smallest units possible but also recognition that data availability, reliability and

confidentiality will always be limiting factors. Obviously, data limitations will depend on the

sources being used and will therefore vary from one indicator to the next but, in general, the

choice of smaller units will limit the range of indicators available.

Second, areas need to be defined in a consistent and fair manner across the country.

Consistency is needed to prevent manipulation of the results in some locations through the

use of highly artificial boundaries designed to pick out pockets of deprivation or the use of

smaller areas. In general, the smaller the area (in population terms), the greater the chance

that it will obtain a very high or very low score on the deprivation index. Larger areas will

tend to score closer to the average. However, the criteria of consistency and fairness need not

51


be taken to imply that the same size of unit have to be applied across the country. It may be

fairer to use smaller units in rural than urban areas to reflect differences in the forms of

development. Urban areas tend to be made up of larger blocks of similar housing so that

larger areas are less likely to contain a mix of deprived and more affluent households. This

was also a view expressed in the neighbourhood definitions work (Flowerdew and Feng,

2002). Several interviewees argued that wards in rural areas were so large as to mask

diversity although this point could equally be made about a great many wards in urban areas.

There was also a desire from some interviewees that units should represent “natural areas” or

“communities” as far as possible but this does not seem to be achievable. It is far from clear

what such terms mean even in theory. They may contain implications of social homogeneity,

social integration or a shared sense of identify but there is no agreement on which of these is

most relevant. In practice, people have different degrees of attachment to or identification

with their locality and different views on how their area should be defined. Devising a set of

boundaries based on notions of “community” therefore proves highly problematic. Recent

work for the Scottish Neighbourhood Statistics project found it impossible to develop

consistent boundaries for “neighbourhoods” or “communities” in this way (Flowerdew and

Feng, 2002). The set of sub-ward boundaries which that work devised were given the more

neutral term, “small data unit”.

In practical terms, the main choice is whether to use wards or these small data units. The

SIMD2003 uses local authority wards and this was welcomed by interviewees as a significant

improvement on postcode sectors. Wards are slightly smaller than sectors (1222 in 1999,

compared with 950 sectors), with an average population of 4100. They nest within local

authority boundaries and seek to take account of natural features. Wards can also be seen to

meet the criteria of fairness. Compared with postcode sectors, they vary much less in size

(within each authority) and tend to be of a more even shape. They also tend to be smaller in

rural authorities than in urban areas (Figure 4.1). In Edinburgh and Glasgow, the average

ward contains 8000 people and in Aberdeen, Dundee and West Lothian it is around 4900. By

contrast, the three Island authorities all have average ward populations below 1000 while

Argyll and Bute, Highland and Clackmannanshire average 2500 to 2700. This is not to argue

that wards represent natural areas or communities or that they capture homogeneous areas in

either rural or urban settings. There is some grounds for arguing, however, that they are at

least “equally unfair” to urban and rural areas. They do have the attraction (for local

authorities at least) of reflecting one of the main concerns of elected representatives. 52

Figure 4.1: Average ward population by Unitary Authority

0

1000

2000

3000

4000

5000

6000

7000

8000

City

of E

dinb

urgh

Gla

sgow

City

Dun

dee

City

Aber

deen

City

Wes

t Lot

hian

Nor

th L

anar

kshi

re

Nor

th A

yrsh

ire

East

Dun

barto

nshi

re

Sout

h La

nark

shire

Mid

loth

ian

Falk

irk Fife

East

Ren

frew

shire

Ren

frew

shire

Wes

t Dun

barto

nshi

re

Inve

rcly

de

East

Lot

hian

Stirl

ing

Sout

h Ay

rshi

re

East

Ayr

shire

Angu

s

Aber

deen

shire

Mor

ay

Perth

& K

inro

ss

Scot

tish

Bord

ers

Dum

fries

& G

allo

way

Cla

ckm

anna

nshi

re

Hig

hlan

d

Argy

ll &

Bute

Shet

land

Isla

nds

Ork

ney

Isla

nds

Com

hairl

e na

n Ei

lean

Sia

r

Ave

rage

war

d po

pula

tion

Source: GRO Scotland, revised mid-year population estimates for 2000.


Small data units would meet the criteria of consistency as they are based on an explicit

methodology set out in Flowerdew and Feng (2002). The starting point for defining them is

the catchment areas of non-denominational primary schools, divided or amalgamated to meet

the target population criteria. Boundaries also seek to reflect natural and environmental

features, and to produce areas of compact shape which are relatively homogeneous in social

terms. They are significantly smaller than wards, with a target population of 500 to 1000

across the country. They would therefore provide much more spatial detail for urban as well

as rural areas but this would inevitably limit the range of indicators which could be used and

increase problems of errors associated with small numbers.

A compromise position would be to adopt a two-tier approach, providing a full index of

deprivation at the higher (ward) level and accompanying this with a reduced set of indicators

or simpler index for the lower (small data unit) level. It might be possible to provide figures

for income deprived or employment deprived for each small data unit, for example, given that

these are constructed from administrative data. This would provide some evidence of the sub-

ward picture. It would highlight, for example, wards which combined small areas with high

concentrations of deprivation alongside areas with little deprivation.

In updating the index in future years, it will be necessary to use “frozen” boundaries. The

boundaries used for the index cannot be changed each time ward or small data unit

boundaries are adjusted. First this would prevent comparison of changes over time. Second,

and more importantly, small changes might lead to the danger of disclosing confidential

information – the “differencing” problem.

Views of the criteria for selecting spatial units and on the relative merits of wards and

small data units would be welcomed, as would alternative suggestions. Views on the

merits of a two-tier approach would also be welcome.

54


4.4 Domains versus life-stage approaches

This section discusses an alternative way in which the indicators used to construct the

SIMD2003, or any future deprivation index, could be combined and presented. The current

index is based around the use of domains such as income, employment or education. Multiple

indicators attempt to capture problems of deprivation within each domain, with domain

scores combined into the final index. It is an explicit feature of this approach that each

domain score has some meaning. For income and employment domains, this is transparent as

the domain score shows the percentage of the population suffering this individual deprivation

(though not the number in multiple deprivation). For the other domains, the domain score

results from combining a number of variables with different scales or referring to different

groups. The education domain, for example, combines exam results for school children,

staying-on rates for those over school leaving age and qualifications held by working adults.

In creating an overall domain score:

“It is hypothesised that an underlying factor exists at an ecological [area] level that

makes these different states likely to exist together in a local area. This underlying

factor cannot be measured directly but can be identified through its effect on

individuals (e.g. failure to obtain qualifications and failure to enter higher education).”

(Noble et al, 2003: p.53: emphasis added)

Factor analysis is the statistical technique used to identify this underlying variable. This

assumes that each indicator measures the underlying problem to a greater or lesser extent, so

that the “best” measure of the underlying factor results from a combination of these, using

weights determined by factor analysis.

Two problems surround this approach. First, methodological questions have been raised over

the use of factor analysis, notably due to the instability of the results produced (see Raab,

2003 for details). A slight change in the selection of indicators can produce a dramatic shift in

the weightings which factor analysis produces and hence in the domain scores. In updating

the index in future years, slight differences in ward scores might produce very different

weightings, making comparisons of change over time problematic. At the very least, there is

a strong argument for re-examining the use of factor analysis and considering the use of

explicit weights (determined with reference to research or public consultation).

55


Second, from a conceptual point of view, it is not clear what processes are thought to

underpin the concentration of unrelated groups with different education (or health or

employment) problems into the same areas. What are the area-level (ecological) factors

which explain why, in some areas, children do badly at school and adults have few

qualifications? The domain approach looks appealing from a policy perspective because it

appears to identify sets of problems which different functional departments can relate to

(health, education, employment). But these domains are not identifying health, education or

employment problems specific to these locations. If there is an underlying explanation for

why different educationally-deprived groups are concentrated into the same areas, it is

connected with the way in which the housing system segregates groups by income (Kleinman

and Whitehead, 1999).

An alternative approach with more obvious meaning would be to measure deprivation for

different life-stage groups – children, young people, working age and older people – and

combine these group deprivation scores to obtain an overall index of deprivation. The same

set of indicators used for the SIMD2003 could be grouped not by domain but by these life-

stage groups. Table 4.1 below shows how this might look. There could be discussion over the

allocation of some indicators (does low birthweight measure child deprivation or the mother’s

deprivation?) and there are some areas which are priorities for being strengthened. With some

additional work, some income and health indicators could be split to give a better breakdown

for young, working age and older people.

In theory, both sets of information could be published in parallel (domain and life-stage

scores), provided this would not compromise confidentiality. This might provide useful

additional detail for little extra work.

Views of the value of the domain and life-stage approaches would be welcomed.

56

Table 4.1: SIMD2003 indicators by domain and life-stage Life stage Income deprivation Employment deprivation Health deprivation and

disability Education, skills and training deprivation

Geographical access to services

Children * Children in IS households * Children in JSA (IB)

households * Children in low income

WFTC households * Children in low income

DTC households

* Proportion of live singleton births of low birth weight (<2,500g)

* Pupil performance on SQA at Stage 4

* Secondary level absences

* Distance to a primary school

Young people (16 to 24)

* Compulsory participants – New Deal for the under 25s

* Young people beyond compulsory school leaving age not in school, FE or HE

Working age (25 to retirement)

* Unemployment claimant count of those aged under 60

* IB and SDA recipients aged under 60

* Compulsory participants – New Deal for 25+

* Working age adults with no qualifications

Post-retirement age

* Adults in IS households * Adults in JSA (IB)

households * Adults in low income

WFTC households * Adults in low income DTC

households

* Comparative Mortality Factor for under 75s

* Hospital episodes related to alcohol use

* Hospital episodes related to drug use

* Comparative Illness Factor (DLA, IB, SDA, AA)

* Emergency admissions to hospital

* Proportion of population prescribed drugs for anxiety, depression or psychosis

* Distance to a GP surgery or health centre

* Distance to a general stores or supermarket

* Distance to a petrol station * Distance to a bank or

building society * Distance to community

internet facilities

Notes: 1. Shaded boxes are not applicable to the relevant group. 2. Some indicators currently in the SIMD2003 cover the whole adult age group, although many could be disaggregated into the individual life-stage groups.


5. Developing measures of individual deprivation

This section considers the third element of our work – measuring deprivation at an individual

rather than area level. In the first part, it considers how individual deprivation might be

measured using a single standard for the whole population. In the second part, it turns to the

question of whether different standards might be developed to reflect the needs of different

social groups in more detail.

Area deprivation indices have been calculated by Government for at least twenty years but

there has been no similar attempt to measure deprivation at an individual level. There are

established measures of the number of individuals in income poverty, showing the number of

people below a given threshold; for the latest analysis for Scotland, see Scottish Executive

(2003). There is an on-going consultation process to devise measures of child poverty (DWP,

2002a). In recent years, the Government has begun to publish a wide range of different

indicators on social exclusion or social injustice, at both UK and Scottish levels (DWP,

2002b; Scottish Executive, 2002b). These provide information on individual aspects of the

problem, however, but do not measure the number of people excluded or suffering injustice.

There is therefore a clear gap in current data and this was recognised by many interviewees.

In particular, there was a concern that the focus on area-based measures led to an over-

emphasis on the problems of a small number of areas and an over-reliance of spatially

targeted programmes as a result.

5.1 Measuring individual deprivation

The approach to measuring deprivation for individuals is quite different to that for small

areas, as noted in Section 3.3 above. To measure deprivation for small areas, area-based

approaches are forced to compromise, relying on a limited number of indicators drawn from

diverse sources. They do not enable us to say how many people are multiply deprived in each

area or to determine the extent to which multiply deprived individuals live in the most

deprived areas. The individual approach uses more direct measures of material and social

deprivation by examining the circumstances of individuals in more detail.

58


One means of achieving this might be by linking the range of information held about each

person by Government in existing databases, as discussed in 3.3. As the feasibility of that

approach is unclear at present, we focus on survey-based approaches in this section. The best

examples in Britain to date have been the Breadline Britain and PSE1999 studies already

mentioned. These involve a three-stage process. In the first stage, a cross-section of the

population is interviewed to determine what the majority consider to be the “necessities of

life” – the material goods or social activities which everybody should be able to enjoy. A

long list of possible items for inclusion is covered. For the PSE1999, the list was drawn up

from an examination of existing research and from in-depth interviews or focus groups

conducted with a range of people living in different situations (Pantazis et al, 2000).

Following the survey, some 35 items were identified as necessities through this process.

Table 5.1 below is taken from that report and shows the proportion of people seeing each

item as a necessity.

Given limitations of cost and practicality, no list can be comprehensive so the focus is on

items which are part of daily life for the great majority of the population. It can be argued that

the list may fail to reflect the needs or situations of particular groups and, as such, would

provide an incomplete measure of deprivation. However, the researchers note that:

“The validity of [this] approach rests on an assumption … that there are not wide

variations in the definition of necessities among different groups of society.

Otherwise, the definition of an unacceptable standard of living just becomes contested

...” (Pantazis et al, 2000: p. 2)

Analysing responses to the first stage survey, the report finds only limited differences

between groups based on gender, age, class, level of education, housing tenure, ethnicity, car

ownership or income. The approach therefore seems justified.

In the second stage, a further survey gathers data from individuals on their income and living

standards to establish the number of people lacking one or more of the 35 necessities as a

result of lack of income. For the PSE1999, interviewees were presented with the list and

asked to say whether they: had the item; did not have it but did not want it; or did not have it

because they could not afford it. The last of these categories identified those deprived through

low income.

59


Table 5.1: Necessities of life in Britain, 1999 Necessary Desirable D/K Beds and bedding for everyone 95 4 Heating to warm living areas 94 5 Damp free home 93 6 1 Visiting friends or family in hospital 92 7 1 Two meals a day 91 9 1 Medicines prescribed by doctor 90 9 1 Refrigerator 89 11 1 Fresh fruit and vegetables daily 86 13 1 A warm waterproof coat 85 14 1 Replace broken electrical goods 85 14 2 Visits to friends or family 84 15 1 Celebrations on special occasions 83 16 2 Money to keep home decorated 82 17 1 Visits to school e.g. sports day 81 17 2 Attending weddings, funerals 80 19 1 Meat, fish or vegetarian equivalent 79 19 1 Insurance of contents of dwelling 79 20 1 A hobby or leisure activity 78 20 1 A washing machine 76 22 1 Collect children from school 75 23 3 Telephone 71 28 1 Appropriate clothes for job interviews 69 28 2 Deep freezer/fridge freezer 68 30 2 Carpets in living rooms and bedrooms 67 31 2 Regular savings for rainy days 66 32 2 Two pairs of all weather shoes 64 34 2 Friends or family round for a meal 64 34 2 Money to spend on self weekly 59 39 2 A television 56 43 2 A roast joint/vegetarian equivalent weekly 56 41 3 Presents for friends/family yearly 56 42 2 A holiday away from home 55 43 3 Replace worn out furniture 54 43 3 A dictionary 53 44 3 An outfit for social occasions 51 46 3 New, not second hand, clothes 48 49 3 Attending place of worship 42 55 4 A car 38 59 3 Coach/train fares to visit friends/family 38 58 4 A evening out once a fortnight 37 56 3 A dressing gown 34 63 4 Having a daily newspaper 30 66 4 A meal in a restaurant/pub monthly 26 71 4 Microwave oven 23 73 4 Tumble dryer 20 75 4 Going to the pub once a fortnight 20 76 4 A video cassette recorder 19 78 3 Holidays abroad once a year 19 77 4 CD player 12 84 4 A home computer 11 85 4 A dishwasher 7 88 5 Mobile phone 7 88 5 Access to the internet 6 89 5 Satellite television 5 90 5

Source: Pantazis et al (2000), Table 1.

60


In the third stage, a cut-off point is applied to determine the number of people considered to

be (multiply) deprived. For the PSE1999, people who had a low income and who lacked two

or more necessities were deemed “poor”. This covered 26 per cent of the population. The

survey also enables comparisons to be made between groups to identify the relative incidence

of deprivation. The poverty rates for unemployed people, lone parents or disabled or long-

term sick people were all over 60 per cent, for example. Comparisons can also be made of

differences between groups in terms of the types of necessity they are missing.

One issue with the survey-based approach is the extent to which it can provide detail on

levels of deprivation for different parts of Scotland. The approach of the PSE1999 was to

cover a relatively small number of individuals in depth. The aim was to explore different

conceptions of poverty, deprivation and social exclusion and to provide greater understanding

of the factors associated these. With a sample of just over 1500 people, it could provide

robust estimates of deprivation for the whole country (Britain) and for different social groups,

but only limited regional breakdown. In the initial interviews for our work, there was strong

interest in being able to get estimates of the number of people deprived down to the level of

local authorities. A dedicated survey to achieve this would be extremely expensive. The

alternative would be build relevant questions into an existing survey. The SHS would be the

most likely option, as it covers many of the necessary topics already. The disadvantage of this

approach is that it limits our ability to gather a wider range of information which might help

to develop our understanding of the factors and processes which tend to lead to deprivation.

Conclusions

From our initial interviews, there was strong support for developing a measure of individual

deprivation, particularly one which would provide estimates down to the level of local

authorities. This would act as a complement to area-based measures. We see the approach

adopted in the PSE1999 as providing the best template for this work as it relies on direct

measures of living standards, tested against a standard determined by a consensual approach.

A similar approach could be implemented in Scotland, most likely by incorporating

additional questions in an existing household survey.

We would welcome feedback on the value of measuring deprivation at an individual

level as well as specific comments on adopting the approach used for the PSE1999 and

on the relative merits of emphasising breadth and spatial detail rather than depth.

61


5.2 Measures for different social groups

A further aim of this work is to explore whether it might be useful and feasible to develop a

range of measures to assess the levels of deprivation for different social groups. As already

noted, the approach underpinning the PSE1999 is to develop a single national standard for

measuring deprivation. This is not to deny that there may be variations between social groups

to some extent but the argument is that there is sufficient commonality to make a single

standard broadly applicable. This approach is chosen partly as a matter of practicality but also

because comparisons between social groups can only be made on the basis of a common

standard.

Some may argue that this common standard is not able to capture the needs of particular

groups adequately. Indeed, this is recognised by the PSE1999 in relation to children where an

entirely separate list of necessities is developed and applied. The question for this project is

whether this approach should be taken further in any Scottish measure of individual

deprivation. In losing the ability to make comparisons between groups, the main gain would

be a better assessment of absolute levels of need for each group and of how the deprived

members of each group were distributed across Scotland.

No group is homogeneous and it will always be possible to propose further sub-divisions

reflecting progressively finer differences in needs and hence ever an increasing range of

deprivation measures. The task for this study is to identify the main groups where there are

sufficient differences to justify the additional costs and complexity.

In the rest of this section, we discuss the arguments for developing separate measures for a

range of social groups. We have chosen to look at life stage groups and gender, ethnicity and

disability. These were the major groups identified in policy documents and were also seen by

interviews as most likely to be of interest. We would welcome comments on whether there is

a need to examine other groups and justifications for this.

We would welcome feedback on the overall value of specifying different standards for

individual social groups. We would also value comments on which groups should be

considered a priority in this respect and why.

62


5.2.1 Children and young people Children and young people are two of the groups identified in the Scottish Executive’s social

justice strategy documents (Scottish Executive, 1999). There are some ambiguities in the

definition of these different groups. Whilst many commentators refer to children as those

aged under sixteen, the social justice framework includes secondary school-aged data in its

young peoples’ strategy.

Children

Children are defined here as being people under the statutory school leaving age (up to and

including sixteen years of age). They comprise a fifth of the Scottish population and are more

likely to be living in poverty than adults (Scottish Executive, 2003). Between 1979 to 1997,

there was a sharp rise in UK child poverty (defined as children living in households with less

than 60 percent of median household income after housing costs) – up from 14 to 34 per cent

(Brewer et al, 2003). Since then, there has been some improvement, with the proportion of

children in Scotland in poverty down from 34 per cent in 1996/7 to 30 per cent in 2001/2 in

both Scotland and the UK as a whole (Scottish Executive, 2003; Brewer et al, 2003).

In parallel with the UK government, the Scottish Executive is committed to eradicating child

poverty in a generation and to ensuring that all children in Scotland achieve an appropriate

level of educational attainment on leaving primary school (Scottish Executive, 2001a;

2002b). The DWP is conducting a formal consultation to establish an agreed measure to track

changes in levels of child poverty in the UK (DWP, 2002a). There are two potential focuses

for public concern: the current welfare of children and their future welfare as adults. For the

purposes of this work, it seems more appropriate to stress current living standards and to

focus on the needs of children as children.

The key issue is the extent to which child deprivation can be determined from examining

household deprivation. As already noted, the PSE1999 contained different standards for child

and adult deprivation. It drew on in-depth interviews with children and young people to

establish a list of child necessities, although the final judgements about what constituted

necessities were made by adult respondents. Nevertheless, one would expect a very high

degree of correlation between child and adult deprivation measures. Although it is known that

some adults in poverty forgo basic necessities in order to minimise the deprivation suffered

by their children (Middleton, 2000), household poverty is generally accepted as the most

important indicator of the well-being of children (Bradshaw, 2002; Bentley et al., 1999).

63


Household poverty and deprivation spill over into other aspects of child deprivation. Children

in low-income households are more likely to live in poor housing (Quilgars and Wallace,

1992). Lower educational achievement remains strongly related to household poverty (Coles

and Kenwright, 2002). There are also linkages between household poverty, truancy and

school exclusion (SEU, 1998). Just under half of all school exclusions in Scotland involve

children who are entitled to free school meals (often taken as a proxy measure of poverty)

(Edwards and Coles, 2002). There are also linkages between poverty and health deprivation,

related to the poor diets of many Scottish children (Searle, 2002) and deprivation is also

related to an increasing risk of physical accidents, most notably links between poor areas and

the number of traffic accidents involving children. As with adults, the social or relational

aspects of child deprivation are less well developed.

Young People

Many of the issues relevant to child deprivation are also applicable to young people (defined

here as 16 to 24 years old). The Executive’s social justice targets are to maximise educational

attainment (skills and qualifications) and to ensure that every 19 year-old is engaged in

education, training or work. Deprivation measures might wish to put greater emphasis on

current circumstances and on social or relational issues.

Low levels of pay and restrictions on access to social housing contribute to the housing

deprivation of young people (Fahmy, 2001). Lower educational achievement is related to

earlier deprivation and exerts a strong influence on immediate labour market outcomes and

future life opportunities for young people (Bentley et al., 1999).

The consultation process identified the desirability of including measures of social processes

relating to levels of deprivation for young people. For example, measures of access to

services might need to capture the discrimination, territorialism and stigmatisation that act as

barriers for young people. This is linked to a broader social aspect of young people’s

deprivation in which the quality off inter-generational relationships (including discrimination,

disaffection and lack of empowerment in decision-making structures) act to deprive young

people of a deeper and more meaningful engagement with their local communities.

64


Conclusions

There appears to be a strong case for developing a separate measure of child deprivation,

given high poverty rates for this group, high policy interest and significant differences in the

items which might be considered necessities. This case would be strengthened by evidence

that child deprivation is not already well captured by household (adult) deprivation measures.

Arguments for separate measures for young people are less clear cut. Any measurement of

deprivation for children or young people would need to be based on the perceptions of these

groups about their own necessities of life.

5.2.2 Older people Older people are defined here as those above retirement age. They represent just over one

fifth of the population in Scotland and this has been rising for some time. They are slightly

more likely to live in rural areas than the average; 38 per cent of the population of remote or

very remote rural areas is above the age of 60, compared with 33 per cent for Scotland as a

whole (Scottish Household Survey data). Until recently, levels of poverty for older people

were above the average but they have now been reduced to the average for the population as

a whole. For Scotland, the proportion of older people in poverty (defined as those of pension

age in households below 60 percent of median household income after housing costs) has

fallen from 26 per cent in 1996/7 to 20 per cent in 2001/2 (Scottish Executive, 2003). Older

people continue to have much higher levels of disability than average (Riddell and Banks,

2001) and this may bring additional problems of poverty and deprivation as discussed in 5.2.5

below.

Older individuals are also far more likely to experience deprivation in relation to their health,

with many suffering from debilitating illnesses or disabilities. This not only deprives them of

full engagement in many aspects of daily life, but also makes them particularly dependent

upon health care and social services provision. Access to such services may be particularly

problematic in rural areas due to the location and costs of providing hospitals, hospices,

community care and medical services. Older people in general, and single older people in

particular, are vulnerable to social isolation and to be excluded from social relations (Patsios,

2000). Such isolation, in addition to physical vulnerability, contributes to the disproportionate

fear of crime experienced by older people in Scotland, although the Scottish crime survey,

which takes place every four years, has shown a reduction between 1993 and 2000 in the

percentage of people aged over 60 worried about crime (Scottish Executive, 2002c)

65


Conclusions The traditional image of old age being associated with poverty has begun to change and older

people are now longer more likely to be in poverty than the average. This is a large group

with high levels of needs, both material and social. Although many of these needs would be

captured by a standard measure of deprivation, there may be a case for separate measures.

5.2.3 Women and men In response to the greater policy emphasis on equality, there has been a call for better

research on the deprivations suffered by women in particular (Scottish Executive, 2000b).

This echoes an earlier research review that found limited data on women in Scotland beyond

the purely descriptive. Small and ad hoc studies did not provide a basis for monitoring

progress with the implementation of an equality strategy or with change in the relative

positions of the two genders arising from social and economic factors. There was little

research providing comparisons between regions within Scotland, or between Scotland and

other parts of the UK.

Measures of deprivation based around a single standard are able to provide a breakdown of

deprivation rates for women and men, as for other social groups, and so provide comparisons

between them. There are concerns, however, that this approach leads to incomplete coverage

of deprivation for both, but particularly for women.

First, there is a danger of making inferences about individual characteristics from household

level data. Household income is not necessarily shared equally between household members

and there is evidence that women do worse on average in this division. Similarly, the

presence of a car for the household does not mean that all adults have equal access to it.

Second, women and men may have different needs or weight problems differently. Almost

four times more women than men fear walking alone in their area after dark (6% of males,

23% of females) (Scottish Executive, 2000c). When homeless, men are much more likely

than women to sleep rough. Women’s homelessness is more likely to take “hidden” forms.

Third, women and men take on different roles in the home and in the labour market on

average. Women’s greater role in caring for children explains their lower labour market

participation rates but also their greater access to social housing (as homelessness legislation

gives priority to households with children).

66


Conclusions Women and men are at risk of being deprived in various gender-specific ways. Public policy

stresses the need to reduce gender-based inequalities. There seems to be a strong case for

developing gender-specific deprivation measures. At minimum, work using a standard

measure of deprivation would need to be complemented by studies to examine the relative

positions of women and men within households.

5.2.4 Black and minority ethnic groups Black and minority ethnic (BME) groups comprise two percent of the Scottish population,

with no single group in the Census classification accounting for more than half this total. This

population is concentrated in Scotland’s largest four cities, with a third living in Glasgow

alone. Scotland also has a small population of asylum-seekers and refugees from very diverse

ethnic backgrounds (Macaskill and Petrie, 2000).

Few large-scale studies have been undertaken into the living standards of BME groups in

Scotland, with most research being confined to small-scale qualitative work (Netto et al.,

2001). Interviews suggested that poverty and deprivation may not be as prevalent among the

BME population in Scotland as in England. The 2001 Census suggests that unemployment

levels are similar for individuals from white, Pakistani and Chinese backgrounds in Scotland,

although levels are higher for other ethnic groups. Other studies in Scotland suggest that

BME groups are more likely to experience unemployment, low-pay and reduced benefit

uptake (Scottish Executive, 2000c). In addition, limited access to the wider labour market

may force many into self-employment, disguising poverty rather than indicating affluence.

This high level of self-employment was confirmed in the 2001 Census. Various factors

including cultural aspects, immigration laws, employment patterns and discrimination have

also been reported to result in non-claiming or under-claming of welfare benefit entitlements,

further contributing to the deprivation of BME groups (Scottish Executive, 2000c).

BME groups are also at risk of experiencing other aspects of deprivation. There is evidence to

suggest that the low take up of public services by BME groups is not due to a lack of need but

rather to a lack of knowledge about the availability of services and a range of institutional or

cultural barriers to accessing them (Netto et al., 2001; Scottish Executive, 2000c). Research

from rural areas suggested additional difficulties for BME groups in accessing services due to

a lack of recognition of their particular needs by service providers in these areas.

67


The experience of many ethnic minorities groups may be conceptualised as one of ‘partial

citizenship’, in which material deprivation is compounded by racism, making it difficult to

participate in society as a full member (Amin and Oppenheim, 1992). There are increasing

incidences of racial harassment in Scotland (Netto et al., 2001). Fear of crime and harassment

impacts on ethnic minority individuals’ choice of housing, and may impact particularly upon

young people in their schools and local environment (Scottish Executive, 2000c; Clark and

Leven, 2002). Although there is no simplistic correlation between discrimination and

deprivation, it may be assumed that discrimination is an important dimension of the

experience of deprivation BME groups in Scotland (Netto et al., 2001).

Conclusions The relatively small size of the BME population would make measuring deprivation by

survey means difficult and the lack of evidence of high levels of deprivation amongst these

groups would also suggest that developing separate measures would be a lower priority. On

the other hand, BME groups may clearly face additional types of problem, including those

arising from problems in accessing some public services and from discrimination.

5.2.5 People with disabilities People with disabilities are a diverse group in terms of needs and this has helped hide the

extent of the problems many face. It has proved difficult to establish the number of disabled

people in Scotland, reflecting a problem arising from varying definitions of disability. One

recent estimate suggests there are 800,000 disabled people in Scotland or a sixth of the

population (Scottish Executive, 1999). The consensus is that this number will increase due to

the ageing population. In terms of particular sub-groups, just under two percent of the

population are estimated to have visual impairment, 18 percent suffer some degree of hearing

loss, and four percent have some form of learning difficulties. Just over fifteen percent of the

Scottish working population has a long term disability substantially affecting their daily

activities.

The connection between disability, poverty and deprivation has been well documented. Using

figures from the SHS, Riddell and Banks (2001) report that a majority of disabled people in

Scotland live in poverty. Disabled people are much more likely to be out of work and

claiming benefits. Other recent studies have affirmed this association between disability,

poverty and deprivation at the UK level (Burchardt, 2000). Burchardt and Zaidi (2003)

estimate the additional costs of living for people with disabilities, highlighting the extent to

68


which standard measures of income fail to measure the real extent of poverty for many

disabled people. There is also evidence that people with disabilities are disadvantaged in

terms of housing, transport and access to services (Riddell and Banks, 2001).

The traditional image of disability has drawn from the medical model which emphasises the

presence of impairment. In this perspective, disability would be seen as a disadvantage in its

own right. This model has been challenged by writers such as Oliver (1990) who argues for

social model of disability. In this conception, people whose abilities differ from the norm are

disabled by both environmental factors (unsuitable housing, inaccessible public buildings)

and social barriers (stigma, discrimination), rather than through having a medical impairment

per se. From this perspective, differences in ability only become disadvantages through social

processes.

Although disabled people may be conceptualised as a distinctive group disadvantaged by a

common set of environmental factors and social barriers (Oliver, 1990), there is still little

knowledge about the actual experiences of disabled people in Scotland (Riddell and Banks,

2001).

Conclusions There appears to be a strong case for considering the development of separate measures of

deprivation for people with disabilities, not least because this is a large group with high levels

of poverty and deprivation. Many disabilities bring higher costs of living but they may also

contribute to poor educational attainment, less access to services, problems of mobility and

social isolation. One challenge would be coping with the diversity of needs of people with

disabilities.

69


5.2.6 Conclusions Some evidence on the relative levels of need between different social groups can be gained

by using a single deprivation measure and comparing levels of deprivation between them.

Indeed, comparability between groups can only be made on the basis of a common standard.

It is therefore essential that any work to measure deprivation using a common standard in

future is able to provide breakdowns of levels of deprivation by life-stage group, gender,

ethnicity and disability.

In some cases, however, there is an argument for developing a separate standard to capture

levels of need within each group more accurately. From this short analysis, the clearest case

would appear to exist in relation to:

• children (on grounds of levels of need and differences in necessities);

• women (on grounds of differences in needs and problems of intra-household resource

distribution);

• disabled people (on grounds of differences of needs, especially cost of living).

The arguments for older people are less clear cut as levels of need are not above average.

With BME groups, there is little evidence of high levels of material need and, at a more

practical level, this is a small group in absolute terms so needs would be difficult to measure.

With young people, the argument for developing a separate standard also appears weaker.

One division not discussed is between sexuality groups, due in large part to lack of evidence

on relative levels of need at present. We would be keen to have comments on whether there is

a need to develop a separate deprivation measure for lesbian, gay, bi-sexual or transgender

people and why. At the very least, work on developing a common standard should ensure that

it is possible to disaggregate levels of deprivation for sexuality groups.

Views of this discussion would be welcome, particularly those which provided

arguments for or against developing deprivation measures for specific groups and

which discussed how these measures might differ from a common standard.

70


6. Updating the measures

Our task is to set out a long-term strategy for developing measures of deprivation in Scotland.

No matter how far-sighted this work tries to be, there will always be the need to review the

measures. There may be opportunities to include new indicators, including those arising from

the development of additional data sources recommended as part of this work. There may be

methodological advances. There will be changes in standards of living and in social

judgements as to what constitute necessities of life. This section briefly considers some of the

issues which arise for updating the measures.

There is a trade-off between retaining a consistent measure to enable analyses to be made of

change over time and updating a measure to take account of new data sources,

methodological advances or changes in social perceptions or practices. Area-deprivation

indices have rarely retained the same method for successive measures, being updated as new

data sources and methods have become available. Up to now, however, these measures have

only been produced on an occasional basis (following the decennial Census, until recently) so

consistency seemed less relevant. The survey-based measures of individual deprivation

developed for the Breadline Britain/PSE1999 work have also changed over time. Following

the logic of a relative definition, they set out to capture changes in perceptions of necessities

(Gordon, Pantazis and Townsend, 2000), but these measures too have been produced at

relatively infrequent intervals (1983, 1990 and 1999).

With the move to producing deprivation measures on a more regular basis, possibly annually,

a compromise position might be useful. This would entail retaining the same measure for a

period of several years. Several revisions could then be made at one time rather than minor

changes made each year. By ensuring overlap in the change years, a more continuous picture

of the changing pattern of deprivation could be developed.

Views are invited on the most appropriate approach to updating the measures of

deprivation in future.

71


7. Conclusions

This report has been produced as part of a project to develop a long-term strategy for

measuring deprivation in Scotland. It is an interim output which sets out our thinking to date.

The aim of the report is to stimulate wider discussion of these issues to inform our final

recommendations. We have tried to set our thinking in a clear and direct manner, and to

accompany this with direct questions. While these are intended to help structure responses,

they are not meant to limit contributions in any sense. We are particularly keen to have

additional research evidence on the nature of deprivation in Scotland brought to our attention.

Inevitably, our work is more advanced in some areas than others at this stage. We have spent

some time setting out a definition and conceptual framework for deprivation, and situating

the term in relation to other concepts of social need. This may seem rather abstract to some,

but it is a necessary stage to go through. We also discuss a number of background issues,

relating to the use of deprivation measures, the Scottish context and the different approaches

to measuring deprivation. With the area deprivation measure, we build on existing work by

Noble et al (2003) in producing the SIMD2003. We have focussed on developing better

coverage by expanding the number of indicators and domains. With the individual

deprivation measure, we suggest building on the model of the PSE1999 survey. We also

begin to explore whether there are particular groups whose needs cannot be adequately

captured by a single standard.

We do not argue that deprivation is in some way a better measure of social needs than any

other concept. No one concept will meet the needs of all situations or potential users. We do

believe, however, that deprivation is important for its focus on the poor material and social

living standards which people face as a result of poverty. In discussing how to measure

deprivation, we have tried to keep to these core issues and not to let the concept expand so far

that this focus is lost. Nevertheless, the overlap between terms such as deprivation and social

exclusion or social justice means that the measurement of deprivation can provide important

insights into a broader conception of social problems and it would be valuable in that respect

as well.

72


References

Amin, K. and Oppenheim, C. (1992) Poverty in black and white: deprivation and ethnic minorities. London: Child Poverty Action Group

Bailey, N., Turok, I. and Docherty, I. (1999) Edinburgh and Glasgow: contrasts in competitiveness and cohesion, Glasgow: Department of Urban Studies.

Beatty, C., Fothergill, S., Gore, T., and Green, A. (2002) The real level of unemployment 2002. Sheffield: Centre for Regional Economic and Social Research, Sheffield Hallam University.

Bentley, T., Oakley, K., Gibson, S. and Kilgour, K. (1999) The real deal: what young people really think about government, politics and social exclusion. London: Demos

Berghman, J. (1995), ‘Social exclusion in Europe: policy context and analytical framework.’ in Room, G. (ed.) (1995), Beyond the threshold: the measurement and analysis of social exclusion’, Bristol: The Policy Press.

Bradshaw, J. (2002) The well-being of children in the UK. London: Save the Children. Bramley, G. and Ford, T. (2001) Social exclusion and lack of access to services: evidence

from the 1999 PSE Survey of Britain. Working paper 14. Bristol: Townsend Centre for International Poverty Research, University of Bristol.

Bramley, G. and Lancaster, S. (1998) Modelling local and small area income distributions in Scotland, Environment and Planning C: Government and Policy 16: 681-706.

Bramley, G., Lancaster, S., and Gordon, D. (2000) Benefit take-up and the geography of poverty in Scotland, Regional Studies 34 (6): 507-19.

Brewer, M., Goodman, A., and Shephard, A. (2003) How has child poverty changed under the Labour government? An update. London: IFS.

Brown, U., Scott, G., Mooney, G. and Duncan, B. (2002) Poverty in Scotland 2002: people, places and policies, Glasgow: CPAG.

Buck, N., Gordon, I., Hall, P., Harloe, M., and Kleinmann, M. (2002) Working capital: life and labour in contemporary London . London: Routledge.

Burchardt, T. (2000) Enduring social exclusion, York: Joseph Rowntree Foundation. Burchardt, T. and Zaidi, A. (2003) Comparing incomes when needs differ: equivalisation for

the extra costs of disability in the UK. CASE Paper 64. London: Centre for the Analysis of Social Exclusion, London School of Economics.

Burns, D., Forrest, R., Flint, J. and Kearns, A. (2001) Empowering Communities: the Impact of Registered Social landlords on Social Capital. Edinburgh: Scottish Homes

Burrows, R. and Rhodes, D. (1998) Unpopular places? Area disadvantage and the geography of misery in England, Bristol: Policy Press.

Carstairs, V. and Morris, R. (1991) Deprivation and health in Scotland. Aberdeen: AUP. 73


Chapman, P., Phimister, E., Shucksmith, M., Upward, R. and Vera-Toscano, E. (1998) Poverty and exclusion in rural Britain: the dynamics of low income and emmployment. York: YPS.

Clark, I. and Leven, T. (2002) 2000 Scottish Crime Survey: analysis of the ethnic minority booster sample. Edinburgh: Scottish Executive Central Research Unit

Coles, B. and Kenwright, H. (2002) Educational Achievement, in J. Bradshaw, (Ed.) The well-being of children in the UK, pp. 231-249. London: Save the Children

Countryside Agency (2003) Updating the English indices of deprivation 2000: Stage 1 consultation report. Response by the Countryside Agency. Cheltenham: Countryside Agency.

Dean, J. and Hastings, A. (2000) Challenging images: housing estates, stigma and regeneration. Bristol: Policy Press.

Department for Work and Pensions (DWP) (2002a) Measuring child poverty: a consultation document. London: DWP.

Department for Work and Pensions (DWP) (2002b) Opportunity for all. Fourth annual report 2002. London: DWP.

Devine, T.M. (ed.) (2000) Scotland’s Shame: bigotry and sectarianism in modern Scotland, Edinburgh: Mainstream.

Duguid, G. (1995) Deprived areas in Scotland. Edinburgh: Scottish Office CRU. Edwards, E. and Coles, B. (2002) Truancy and School Exclusions, in J. Bradshaw, (Ed.) The

Well-being of Children in the UK, pp. 271-287. London: Save the Children Edinburgh Capital City Partnership (1996) Closing the gap. Edinburgh: ECCP. Etzioni, A. (1995) The spirit of community: rights, responsibilities and the communitarian

agenda. London: Fontana. Fahmy, E. (2001) Youth, poverty and social exclusion. PSE1999, Working paper 27. Bristol:

Townsend Centre for International Poverty Research, University of Bristol. Flowerdew, R. and Feng, Z. (2002) Scottish Neighbourhood Statistics neighbourhood

definitions project: phase 2, stage C and D. St Andrews: School of Geography and Geosciences, University of St Andrews.

Forrest, R. and Kearns, A. (2001) Social cohesion, social capital and the neighbourhood, Urban Studies, 38(12): 2125-2144

Gibb, K., Kearns, A., Keoghan, M., Mackay, D., and Turok, I. (1998) Revising the Scottish area deprivation index: volume 1. Edinburgh: SO CRU.

Goodwin, D., Adelman, L., Middleton, S., and Ashworth, K. (1999) Debt, money management and access to financial services: evidence from the 1999 PSE Survey of Britain. Working Paper 8. Bristol: Townsend Centre for International Poverty Research, University of Bristol.

74


Gordon, D., Adelman, L., Ashworth, K., Bradshaw, J., Levitas, R., Middleton, S., Pantazis, C., Patsios, D., Payne, S., Townsend, P., and Williams, J. (2000) Poverty and social exclusion in Britain: report of the Poverty and Social Exclusion Survey of Britain. York: JRF.

Gordon, D. and Pantazis, C. (eds.) (1997) Breadline Britain in the 1990s, Aldershot: Ashgate. Gordon, D., Pantazis, C., and Townsend, P. (2000) Changing necessities of life 1983-99.

Working Paper 3. Bristol: Townsend Centre for International Poverty Research, University of Bristol.

Granovetter, M. (1973) The strength of weak ties, American Journal of Sociology 78 (6): 1360-80.

Harrop, A. and Palmer, G. (2002) Indicators of poverty and social exclusion in rural England 2002: a report for the Countryside Agency. London: New Policy Institute.

Hills, J., Le Grand, J. and Piachaud, D. (eds.) (2002) Understanding social exclusion, Oxford: OUP.

HM Treasury (1999) PAT 14: Access to financial services. London: HM Treasury. Housing Improvement Task Force (2003) Stewardship and responsibility: a policy

framework for private housing in Scotland. Final report. Edinburgh: Scottish Executive.

Ihlanfeldt, K. R. and Sjoquist, D. L. (1998) The spatial mismatch hypothesis: a review of recent studies and their implications for welfare reform, Housing Policy Debate 9 (4): 849-92.

Immergluck, D. (1998) Job proximity and the urban employment problem: do suitable nearby jobs reduce neighbourhood employment rates?, Urban Studies 35 (1): 7-23.

Kearns, A. and Forrest, R. (2000) Social cohesion and multi-level governance, Urban Studies, 37(5), pp. 995-1107.

Kempson, E. and White, M. (2001) Ways in and out of low income in rural England. Cheltenham: Countryside Agency. [unpublished report]

Kleinman, M. and Whitehead, C. (1999) Housing and regeneration: the problem or the solution, National Institute Economic Review (170): 78-86.

Levitas, R (1998) The inclusive society? Social exclusion and New Labour, Basingstoke: Macmillan.

Macaskill, S. and Petrie, M. (2000) I didn’t come here for fun: listening to the views of children and young people who are refugees or asylum seekers in Scotland. Edinburgh: Save the Children

Mack, J. and Lansley, S. (1985) Poor Britain, London: Allen and Unwin. Middleton, S. (2000) Child poverty and social exclusion. PSE1999, Working paper 19.

Bristol: Townsend Centre for International Poverty Research, University of Bristol.

75


Midgley, J. (2003) Access to financial services: moving beyond the spatial perspective, European Research in Regional Science 12 : 162-73.

Midwinter, A. and Monaghan, C. (1990) The measurement and analysis of rural disadvantage. Edinburgh: COSLA.

Miller, D. (1999) Principles of social justice, London: Harvard University Press. Monk, S., Hodge, I., and Midgley, J. (2003) Statistical indicators of rural disadvantage: a

comparison between the index of multiple deprivation and the 'bundles' approach, European Research in Regional Science 12: 76-94.

Netto, G., Arshad, R., deLima, P., Diniz, F., MacEwen, M., Patel, V. and Syed, R. (2001) Audit of Research on Minority Ethnic Issues in Scotland from a ‘Race’ Perspective. Edinburgh: Scottish Executive Central Research Unit

Nolan, B. and Whelan, C. (1996) Resources, Deprivation and Poverty, Oxford: Clarendon Press.

Noble, M., Smith, G. A. N., Penhale, B., Wright, G., Dibben, C., Owen, T., and Lloyd, M. (2000) Measuring multiple deprivation at the small area level: the Indices of Deprivation 2000 . London: DETR.

Noble, M., Smith, G. A. N., Wright, G., Dibben, C., Lloyd, M., and Penhale, B. (2000) Welsh index of multiple deprivation. London: National Statistics.

Noble, M., Smith, G. A. N., Wright, G., Dibben, C., Lloyd, M., and Penhale, B. (2001) Northern Ireland multiple deprivation measure 2001. Belfast: NISRA.

Noble, M. et al. (2002) Updating the English Indices of Deprivation 2000: stage 1 consultation report. London: NRU, ODPM.

Noble, M. Smith, G., Wright, G., Dibben, C. and Lloyd, M., Ratcliffe, A., McLellan, D., Sigala, M. and Anttila, C. (2003) Scottish Indices of Deprivation 2003, Edinburgh: Scottish Executive.

Nozick, R. (1975) Anarchy, state and utopia. Oxford: Blackwell. Oliver, M. (1990) The Politics of Disablement, London: Basingstoke. Pantazis, C., Gordon, D. and Townsend, P. (2000) Perceptions of the necessities of life:

Scotland and England compared. Working paper 24, Bristol: Townsend Centre for International Poverty Research, University of Bristol.

Parkes, A., Kearns, A., and Atkinson, R. (2002) The determinants of neighbourhood dissatisfaction. Glasgow: Centre for Neighbourhood Research, University of Glasgow.

Patsios, D. (2000) Poverty and social exclusion amongst the elderly. PSE1999, Working paper 20. Bristol: Townsend Centre for International Poverty Research, University of Bristol.

76


Phillips, A. (2000) Equality, pluralism, universality: current concerns in normative theory, British Journal of Politics and International Relations 2 (2): 237-55.

Piachaud, D. (1987) ‘Problems in the definition and measurement of poverty’, Journal of Social Policy, 16 (2), pp. 125-146.

Raab, G. (2003) Scottish Index of Multiple Deprivaiton 2003: quality assurance of data and statistical procedures. Unpublished report to Scottish Executive. [Available from Central Statistical Unit at the Scottish Executive]

Riddell, S. and Banks, P. (2001) Disability in Scotland: a baseline study, Glasgow: Strathclyde Centre for Disability Research.

Ringen, S. (1988) Direct and indirect measures of poverty’, Journal of Social Policy,17(3): 351-365.

Robson, B., Deas, I., Mazzei, M., Whyte, D., Hirschfield, A., Bowers, K. and Johnson, S. (2002) Developing new approaches to the measurement of deprivation, London: Association of London Government.

Room, G. (ed.) (1995) Beyond the threshold: the measurement and analysis of social exclusion, Bristol: Policy Press.

Rural Poverty and Inclusion Working Group (RPIWG) (2001) Poverty and social exclusion in rural Scotland. Edinburgh: Scottish Executive.

Scottish Executive (1999) Social justice: a Scotland where everyone matters, Edinburgh: Scottish Executive.

Scottish Executive (2000a) Scottish Health Survey 1998. Edinburgh: Scottish Executive. Scottish Executive (2000b) Equality in Scotland - women and men, Edinburgh: Scottish

Executive. Scottish Executive (2000c) Equality strategy: working together for equality, Edinburgh:

Scottish Executive. Scottish Executive (2001a) Adult literacy and numeracy in Scotland. Edinburgh: Scottish

Executive. Scottish Executive (2001b) Digital inclusion: connecting Scotland's people. Edinburgh:

Scottish Executive. Scottish Executive (2002a) Better communities in Scotland: Closing the gap, Edinburgh:

Scottish Executive. Scottish Executive (2002b) Social justice: a Scotland where everyone matters. Annual report

2002 – technical volume. Edinburgh: Scottish Executive. Scottish Executive (2002c) 2000 Scottish Crime Survey: overview report. Edinburgh:

Scottish Executive Central Research Unit

77


78

Scottish Executive (2003) Households Below Average Income (HBAI) analysis 2001/02: figures for Scotland using the range of low income thresholds, 1994/5 - 2001/02. Edinburgh: Scottish Executive.

Searle, B. (2002) Youth Suicide, in J. Bradshaw, (Ed.) The well-being of children in the UK, pp.320-330. London: Save the Children

Shaw, M., Dorling, D., Gordon, D. and Davey Smith, G. (1999) The widening gap: health inequalities and policy in Britain, Bristol: Policy Press.

Shen, Q. (1998) Location characteristics of inner-city neighbourhoods and employment accessibility of low-wage workers, Environment and Planning B: Planning and Design 25: 345-65.

Shucksmith, M. (1990) The definition of rural areas and rural deprivation. Edinburgh: Scottish Homes, Research Report No.2.

Shucksmith, M., Chapman, P., Clark, G., Black, S., and Conway, E. (1994) Disadvantage in rural Scotland: how is it experienced and how can it be tackled? Glasgow: Scottish Consumer Council

Shucksmith, M., Roberts, D., Scott, D., Chapman, P., and Conway, E. (1996) Disadvantage in rural areas. Research Report 29. London: Rural Development Commission.

Shucksmith, M. & Philip, L. (2000), Social Exclusion in Rural Areas: A Literature Review and Conceptual Framework, Edinburgh: Scottish Executive Central Research Unit.

Social Exclusion Unit (SEU) (1998) Truancy and school exclusion. London: Stationery Office.

Townsend, P. (1987) Deprivation, Journal of Social Policy, 16 (1), pp. 125-146. Webster, D. (2000) The geographical concentration of labour market disadvantage, Oxford

Review of Economic Policy 16 (1): 114-28. Young, I.M. (1990) Justice and the politics of difference, Princeton, NJ.: Princeton University

Press.

Measuring deprivation in Scotland: developing a long-term strategy: interim report

Documents