A study in Birmingham using data from January 1998 by Lucinda Platt and Michael Noble at the Department of Applied Social Studies and Social Research, University of Oxford, shows great diversity in the experience of those on low income according to their ethnic group. The study defined low income as being in receipt of the means-tested Housing Benefit and/or Council Tax Benefit (HB/CTB) and concentrated on four groups: white UK, Bangladeshi, Black Caribbean and Pakistani. It found:

• Comparison with the Census population shows that the Bangladeshi, Black Caribbean, and Pakistani ethnic groups are over-represented in the low-income population and the white population is slightly under-represented.

• Within the low-income population:

- Fifty-five per cent of Bangladeshi families consist of a couple with children, whereas just 7.5 per cent of white UK headed families consist of a couple with children.

- Fifty-five per cent of white UK families contain a person aged 60 or over and nearly half of white UK people aged 60 and over are in receipt of HB/CTB.

- Only 29 per cent of Pakistani and Bangladeshi families contain a person aged 60 or over; yet nearly all Pakistani and Bangladeshi people aged 60 and over are living on Income Support.

- Lone parents make up a larger proportion of Black Caribbean families than they do of any other group. Yet these Black Caribbean lone parents are significantly more likely to be working than the lone parents from any other group.

- Bangladeshi and Pakistani lone parents have more children on average than white UK or Black Caribbean lone parents; yet the average age of Pakistani and Bangladeshi lone parents is up to 10 years older than the other two groups.

- Tenure patterns also vary by ethnic group. Pakistani families are most likely to be owner-occupiers with white UK families most likely to be living in social housing.

A large body of research has demonstrated that, on average, people from minority ethnic groups experience a substantial degree of economic disadvantage, but there has been little work on the specific nature of the experience of low income among different ethnic groups. This study explores differences in the experience of poverty using ethnically coded administrative data.

The project investigated low income (defined as receipt of Housing Benefit or Council Tax Benefit (HB/CTB)) in Birmingham – a city of a million inhabitants – at the beginning of 1998. Birmingham defines 12 ethnic groups; this study concentrates on Bangladeshi, Pakistani, white UK and an aggregate of two of the ‘Black’ groups, called here ‘Black Caribbean’.

At this date, 27 per cent of the city population were on low incomes; but some age groups were over-represented within this total (see Figure 1). Over 30 per cent of both children and pensioners were living in families supported by HB/CTB; whereas the proportion of other age ranges on benefit was lower. Children and older people were thus at greatest risk of poverty, and the percentage in poverty rose markedly after the age of 50.

The study illustrates how the low-income population varies in three different ways according to the ethnic group of the benefit claimant:

• some groups are over- or under-represented in the low-income population;
• the composition of the low-income population varies by ethnic group (i.e. the share of different ages or different family types varies for different groups);
• particular family types show different characteristics according to ethnic group.

Representation in low-income population
Birmingham has a far more diverse ethnic profile than the country as a whole with about 22 per cent of its population from minority groups. Within its low-income population there is an even higher representation of people from minority groups, demonstrating their over-representation amongst those in poverty. Figure 2 compares the breakdown of Birmingham’s overall population by ethnic group (according to the 1991 Census) with those receiving Housing or Council Tax Benefit (where they can be broken down by the same ethnic groups). It illustrates the representation of the different groups amongst those on a low income. Particularly noticeable is the marked over-representation in the low-income population of the Pakistani and Bangladeshi ethnic groups.

Family composition
There are also marked differences in the family composition of the low-income population between ethnic groups. Figure 3 shows the family type composition of the low-income populations of four groups and compares them with the whole dataset. The share of the Bangladeshi low-income population made up of couples with children (55 per cent) is particularly striking, as is the share of lone parents in the Black Caribbean low-income population (34 per cent).

Table 1 illustrates different characteristics of one particular family type, lone parents, according to ethnic group. Black Caribbean lone parents on low income are more likely to be working (‘Non-IS’ cases) than lone parents from other groups. Bangladeshi lone parents are, on average, considerably older than other lone parents.

Housing tenure
The tenure patterns of those on low income also vary by ethnic group:

• Pakistani families are much more likely to own their own homes than other groups.
• White UK families have the highest proportion living in local authority/housing association housing of all groups and the smallest proportion living as private tenants.
• Bangladeshi families have tenure patterns closer to the white UK families than to Pakistani families.

Conclusion
The researchers conclude that the policy implications of the study are threefold:

• For a range of historical and demographic reasons, the use of means-tested benefits by those from Bangladeshi and Pakistani minority groups is particularly extensive. Yet the rules which govern the administration of benefits mean that those at greatest risk of poverty are actually getting equivalently less value out of the benefits. There is a need for a reconsideration of how income maintenance policies can more appropriately, sensitively and equitably support those people from different ethnic groups who are currently living in poverty.
• There is great diversity within the low-income population and sub-groups of it (e.g. lone parents). Most policy initiatives are based on aggregate benefit information; yet this study shows that such aggregates are merely the composite of several diverse populations with their own distinct profiles and trajectories, for each of which the policies may not be entirely appropriate.
• Targeted initiatives for particular sub-groups (such as the New Deal for Lone Parents) should not lose sight of the fact that they may be inappropriate for significant sections of the targeted population and for particular ethnic groups, where, for example, the lone parents may tend to be older and lacking in labour market experience.

In addition, the study demonstrates that administrative records with an ethnic minority data ‘field’ can be analysed imaginatively to look at broad equity issues as well as benefit specific issues.

About the study
The study was based on an extract of all Housing Benefit and Council Tax Benefit cases taken on 14 January 1998. The study used the full extract representing over 140,000 live cases (i.e. they were in receipt of benefit at that point) containing over 270,000 people. Ethnic group was requested on the HB/CTB form, but the completion of the field within the data was insufficient for the study’s purposes and so was supplemented by manually checking responses on a random sample of cases and then applying weights. The application of weights provided a positive ethnic group response for 81 per cent of the data which was then used for analysis. Cases where there was non-response to the ethnic group question showed a slight over-representation of young single people living in privately rented accommodation.

This title is now out of print.