英國dissertation網之英國人力招聘監測數據案例分析,Equal opportunities and recruitment
How Census data can help employers to assess their practices
Shirley Dex and Kingsley Purdam
Contents
Acknowledgements vi
1 Introduction 1
Policy context 1
Why monitor? 4
Is monitoring really necessary? 8
Plan of this report 10
2 Employer case studies 11
Introduction 11
Content of interviews 12
Key findings 12
Comparisons with company regulation in the USA 22
Conclusions 26
3 Available data 28
Introduction 28
Demographic characteristics required 29
Access to the data sources 30
Comparisons with the USA Census data 32
Conclusions 35
4 英國人力資源Case study illustrations of monitoring statistics in the UK 36
Introduction 36
Illustrations 36
Statistical tools for assessing the significance of differences in recruitment 53
5 Discussion and conclusions 55
Main findings 55
Recommendations 56
Best practice 58
Voluntarism or legal requirements 59
Notes 62
References 63
Appendix 1: Employer interviews – topic guide 67
Appendix 2: Questions and codes of key variables in datasets 69
Appendix 3: Supplementary tables associated with figures in Chapter 4 79
Appendix 4: How to tell if statistics imply discrimination 82Acknowledgements
This project could not have been carried out without the assistance and advice of alarge number of individuals and organisations. First, we would like to thank Professor
Angela Dale who provided invaluable help and advice. We thank the JosephRowntree Foundation for providing the funding. We wish to thank the ten UKemployers who agreed to give up valuable time to tell us about their equalopportunities policies and practices. We also thank the four USA employers whospent time telling us how equal opportunities legislation affected their practices and
how they used external Census data to help them comply with their legal duties.We thank the members of our Advisory Group who offered valuable advice andfeedback on the scope and content of the study as well as a draft of this report. Inparticular we thank: Jo Morris (Trades Union Congress), Mary Coussey (University ofCambridge), Gabrielle Cox and Akhtar Zaman (North West Development Agency),Carolyn Davies (Department of Health), Joy Dobbs (Office for National Statistics),Deirdre Golden (Organisation Resource Counselors Inc.), Stephen Williams (ACAS),
John Sharman (Equal Opportunities Commission), Ludi Simpson (University ofManchester) and Vas Patel (Commission for Racial Equality).
Two organisations were extremely helpful in offering introductions to potential employerswho could be case studies for this project – ORC (Organisation Resource CounselorsInc.) and the Equal Opportunities Commission. We also thank Anne Harrop and Chris#p#分頁標題#e#
Goulden of the Joseph Rowntree Foundation for their research management of theproject. Leslie Hathaway of EEO Visions was very helpful in spending time telling usabout USA software developments aimed at helping organisations comply with legalrequirements on equal opportunities monitoring in the USA.
The permission of the Office for National Statistics to use the Samples ofAnonymised Records is gratefully acknowledged, as is the help of Jo Watham at the
Centre for Census and Survey Research, at the University of Manchester supportedby the ESRC/JISC Census of Population Programme.
We also wish to acknowledge the help in obtaining relevant data on the Labour Force Survey (LFS) from Joanne Lindley at the Centre for Census and Survey
Research at the University of Manchester and the ESRC Data Archive for permissionto use the LFS data in this project.
The Census output is Crown copyright and is reproduced with the permission of theController of HMSO and the Queen’s Printer for Scotland.
The authors alone are responsible for the interpretation of the data.
vi1
1 Introduction
Data that would allow employers to give serious examination as to whether theirrecruitment practices might be discriminatory, unlike in the USA, has not been readilyavailable in the UK. Employers have not been able to know the likely pools ofsuitably qualified applicants in the relevant spatial labour market. The availability ofthe 2001 Census data, with its extended range of questions, created a newopportunity for employers to be provided with labour market data, in comparison withwhich they could assess their own statistics. This project set out to:
1 help demonstrate how equal opportunities policies can be implemented moreeffectively in organisations’ recruitment and selection practices
2 devise and demonstrate a method for ensuring that organisations in the UK arenot being discriminatory in their recruitment and hiring decisions
3 help public organisations face their new duties under the Race Relations
Amendment Act.
We hoped to cover a number of population groups in this consideration, notablyminority ethnic groups and gender groups. In addition, we intended to explorewhether the methodology would be appropriate for disabled people and age groups.
Policy context
It is now widely accepted in British society, and embedded in its laws, thatdiscrimination in the workplace and in recruitment to jobs is unacceptable. Since1975, it has been unlawful for employers in Britain to discriminate directly orindirectly on racial grounds and on grounds of sex and marital status in recruitmentand promotion decisions, as well as in advertisements for job vacancies. Since 1996,legislation has also prohibited discrimination against disabled people and has placedresponsibilities on employers to make jobs and facilities accessible to disabledpeople (Disability Discrimination Acts 1995 and2005). The Race Relations
Amendment Act, which came into force in January 2001, takes the responsibilities ofpublic sector bodies further in this area. A range of other national initiatives havebeen drawn up and these complement the legislation (Box 1). In December 2003,under the European Union’s common framework to tackle unfair discrimination, theUK introduced new legislation to outlaw discrimination at work and in training on the11#p#分頁標題#e#
2 Employer case studies
Introduction
The first stage of the research focused on gathering information on the types ofworkforce data and databases collected and maintained by a number of case studyemployers. We felt we needed a good understanding of a range of employers’recruitment practices, job specifications and workforce characteristics in order to see how far available data could help them in their monitoring processes.It was recognised that, ultimately, the outputs from this project would aim to reach allemployers, both small and large, advanced and backward, in implementing andmonitoring equal opportunities. However, to illustrate the full extent of what could bedone with population data sources, we needed to recruit employers who werereasonably well advanced in their commitment to equal opportunities, and in datacollection and collation of relevant applicants’ and workforce data. This meantrecruiting both reasonably large employers who were committed to equalopportunities and public sector employers who were all obliged to act on statutoryrequirements. The organisations were selected through liaison with contactsprovided by the Advisory Group for this project, local knowledge, employersidentified in relevant newsletters and a presentation by the research team to a groupof potentially appropriate employers. It was surprising to find that, even among thesepurposively selected case study employers, there were serious limitations in thescope and accuracy of employers’ data and in job definitions.Interviews were initially conducted with nine employers’ representatives over theautumn of 2001 into 2002. One further employer was recruited later in the project. Intotal, five public sector organisations, four private sector organisations and one voluntary sector organisation were recruited. The employers’ representativesinterviewed were either human resources directors or personnel officers who wereresponsible for equal opportunities, orspecialist equal opportunities directors. Allorganisations were enthusiastic about the research and keen to participate further.
In addition, discussions were held and help for the project was given by representatives of the Commission for Racial Equality and the Equal OpportunitiesCommission. Trade union representatives in each organisation were also informedabout the research.28
3 Available data
Introduction
In this chapter, we examine the datasets available in the UK and their potential toassist employers in monitoring equal opportunities in their workforce. We consider
two datasets only – the 2001 UK Census and the Labour Force Survey. These arethe only datasets likely to offer samples of sufficient size for analyses of minority
ethnic groups as well as containing other necessary information. (See Boxes 12 and13 for a brief description of the two datasets.)
Box 12 2001 UK Census data
The UK Census data for 2001 is the most comprehensive data on the populationresident in the UK. It is certainly the largest data source on minority ethnic#p#分頁標題#e#
populations in the UK. Moreover, categories for collecting data about minorityethnic identity are devised for the Census and tend to influence categories usedin all other official datasets. The 2001 Census obtained responses from 94 percent of the population. Non-response was estimated to be higher for young menand those living in urban areas.
Box 13 The quarterly Labour Force Survey
The quarterly Labour Force Survey is a rolling household-based sample surveyof individuals across the UK focusing on participation in the labour market.Individuals stay in the survey for five successive quarters and are dropped on arotating basis. The LFS collects core information on: industry sectors, type ofwork and occupation; earnings; qualification levels; economic activity; alongside
other personal characteristics (age, gender, disability status, ethnic identity). Inparticular years, additional questions are included. In each quarter, interviewsare achieved at approximately 59,000 addresses with 138,000 individuals.Although it is a very large survey, individuals from particular minority ethnicgroups are low in number in any one year’s sample. Larger groupings sufficientto carry out some analyses are possible by pooling several years of data.
Analyses at levels of geography below region also run into problems of smallsample sizes. There is also an annual LFS (LLFS) consisting of quarterly LFS(Continued)36
4 Case study illustrations of
monitoring statistics in the UK
Introduction
Below we offer a series of illustrations from our organisation case studies of how UKCensus data could be used in the monitoring of organisations’ equal opportunitiespolicies. These cases have been selected in order to provide illustrations acrossdifferent types of equality issue, a range of vacancies, job criteria and differentspatial areas over which recruitment takes place.Illustrations
Illustration 1: Energise
Case of technical staff (mainly professional and science engineers) and gender
applicant profile.
Background
The organisation is an international manufacturer, which has over 100 centresworldwide. Because of the associated nature of some of its businesses, it was notable to give figures for the total employees across the global business.Workforce composition 2001–02
http://www.400190.com/Thesis_Writing/HRM/The organisation was in the process of creating a central database and reviewing allits recruitment methodologies.55
5 Discussion and conclusions
Main findings
This project has highlighted a number of conclusions about equal opportunities andtheir implementation by UK employers. Many organisations have adopted equalopportunities policies but have not moved forward in their monitoring to even a basiclevel, either because they do not collect necessary data or because they do notanalyse the data they collect on their workforce and applicants.
After 30 years of legislation, many employers still do not have a profile of theirworkforce and job applicants broken down by gender or race. The employers we#p#分頁標題#e#
examined were all large employers, all considering themselves to have acommitment to being equal opportunities employers. In this sense, our findingsagree with those of Hoque and Noon (2001) that fewemployers have put systems inplace to ensure equal opportunities in practice, although they claim equalopportunities in written statements and advertisements. There was evidence thatpublic sector organisations were moving towards the goal of better data collectionand analysis of their workforce as part of equal opportunity monitoring at the time ofour fieldwork interviews. Data collection and analysis of applicant data was still weakin a number of these public sector organisations. Private sector organisations were
on the whole well behind those in the public sector in collecting data on theirworkforce and applicants.
All employers were faced with problems of incomplete data, especially on classifyingemployees’ ethnic origin and disability. None had complete data and none had clearstrategies for dealing with the incompleteness of their employee or applicant data.
The employers studied did not commit anything approaching the level of humanresources to data collection or monitoring that is standard in similar organisations inthe USA.
The 2001 UK Census data could be useful in monitoring both employee profiles andapplicants’ data, especially for discrimination by gender or ethnic origin. This projectillustrated the potential of this new data through a number of illustrations. However,there were problems in accessing the data, which would need to be overcome beforethis could be a useful tool for employers. On the basis of current access to 2001Census data, there are many barriers to employers using the data in assisting theirEO monitoring. Employers will need clear guidance on how to monitor effectively aswell as access to clear and easy-to-read tables in order to make comparisons with62
Notes
Chapter 1
1 The research found considerable variation by size of organisation and sector: jobtitle/grade analysis by ethnic group (total 22 per cent; small and medium-sizedenterprises [SME] 18 per cent; large 35 per cent; industrial 25 per cent; services20 per cent); job title/grade analysis by gender (total 31 per cent; SME 26 percent; large 47 per cent; industrial 31 per cent; services 31 per cent).
2 Hoque and Noon (2001) used the 1998 Workplace and Employment RelationsSurvey conducted by the Department of Trade and Industry.
Chapter 2
1 It might be possible to check the accuracy of this assumption using Health
Survey data.
2 One EO manager is quoted as saying: ‘Preparing affirmative action plans was a
defining moment’.
Chapter 3
1 In June 2005, a new survey, the Annual Population Survey (APS), was published.
This is based on the LLFS incorporating a further boost sample. It may offerbetter opportunities for analyses than have previously been possible.
2 We have been advised that Government Register Office Scotland (GROS) did nothave a delay of this length.#p#分頁標題#e#
Chapter 4
1 It is possible that some of the applications are from the same individuals.63
References
Berthoud, R. (2000) ‘Ethnic employment penalties in Britain’, Journal of Ethnic and
Migration Studies, Vol. 26, No. 3
Bhavnani, R. (1994) Black Women in the Labour Market: A Research Review.
Manchester: Equal Opportunities Commission
Brown, C. and Gay, P. (1985) Racial Discrimination: 17 Years after the Act. London:
PSI
Burchardt, T. (2000) Enduring Economic Exclusion: Disabled People, Income and
Work. York: York Publishing Services/Joseph Rowntree Foundation
Burchardt, T. (2004) Being and Becoming: Social Exclusion and the Onset of
Disability. London: Centre for Analysis of Social Exclusion, London School of
Economics
Cabinet Office (2003) Ethnic Minorities and the Labour Market. London: Cabinet
Office
CRE (Commission for Racial Equality) (2001a) The Duty to Promote Race Equality:
A Guide for Public Authorities. London: Commission for Racial Equality
CRE (2001b) Code of Practice on the Duty to Promote Race Equality. London:
Commission for Racial Equality
CRE (2002a) Ethnic monitoring – A Guide for Public Authorities. London:
Commission for Racial Equality
CRE (2002b) Public Procurement and Race Equality. London: Commission for
Racial Equality
CRE (2003a) Public Authorities and Partnerships: A Guide to the Duty to Promote
Race Equality. London: Commission for Racial Equality
CRE (2003b) Formal Investigation Reports – Employment. London: Commission for
Racial Equality
CRE (2003c) Towards Racial Equality. London: Commission for Racial Equality67
Appendix 1: Employer interviews –
topic guide
Background: pre-interview preparation
Nature of business, size, main sites, workforce, recruitment practices, equal
opportunities policies.
Screening over telephone
What data are collected about applicants for your different jobs?
We are looking for at least some employers who collect/record and store data about
applicants for jobs, which includes: minority group, gender.
Main topics to cover
1 Background: confirm nature of business and types of jobs covered in business.
Number of sites. If possible get a profile of workforce; numbers in different jobs
(ask in advance). Ask about minorities, gender, age and disability in each
groups? Is it known? Can you see the numbers? If public sector , get insight into
how they are interpreting their new duties under the Race Relations Amendment
Act to collect data and monitor position of ethnic minority groups (either here or
under EO below).
2 Talk through recruitment process from deciding need to employ/replace staff
member to getting someone in post. Do by a number of jobs of different levels.
Any outsourcing? (Check later about how this affects EO policies/data collected#p#分頁標題#e#
about applicants.)
3 Focus on criteria used in recruiting staff to (range of) different jobs. Are these set
out in writing anywhere? How far do they correspond with educational levels/
qualifications? Role of experience? Age?69
Table A2.1 Variables of interest in the 2001 Census and Labour Force Survey
(2001 onwards)
2001 Census 2003 LFS
Ethnicity
White Same as 2001 Census
British
Irish
Other white
Mixed
White and black Caribbean
White and black African
White and Asian
Other mixed
Asian or Asian British
Indian
Pakistani
Bangladeshi
Other Asian
Black or black British
Caribbean
African
Other black
Chinese or other ethnic group
Chinese
Other ethnic group
Disability
1 Limiting long-term illness or disability 1 DDA disabled and work-limiting disabled
2 Economic activity (last week) DDA disabled
Permanently sick or disabled Work-limiting disabled only
Not disabled
2 Current disability only
Current and past disabled
Past disability only
Not DDA disabled
Appendix 2: Questions and codes of
key variables in datasets
The variables of interest in the 2001 Census and Labour Force Survey (2001
onwards) are defined as shown in Table A2.1.
(Continued)79
Appendix 3: Supplementary tables
associated with figures in Chapter 4
Table A3.1 Energise technical staff recruitment figures 2003–04, and 2001 and
SAR 2001 Census data for given populations
Energise 2001 SAR 2001 SAR Census, 2001 SAR 2001 SAR
‘technical’ Census, EWN, EWN, NVQ Census, EWN, Census, NW,
category NVQ 4/5, 4/5+ technical*, technical*, technical*,
2003–04 20–59 20–59 20–59 20–59
Applicants
Men 77.9 50.1 82.1 81.6 83.2
Women 22.1 49.9 17.9 18.4 16.8
Total % 100 100 100 100 100
N 12,121 205,665 10,177 22,924 2,939
Hires
Men 73.6
Women 26.4
Total % 100
N 197
% applicant men
hired 1.5 (n = 9,437)
% applicant
women hired 1.9 (n = 2,684)
Source: SAR 2001 individual SAR 3 per cent sample England, Wales or NI. Qualification levels differ
in Scotland and are provided separately in the SAR data and not included here. Age 20–59.
EWN: England, Wales or Northern Ireland
NW: North West region only.
NVQ 4/5: highest qualification levels NVQ levels 4 or 5.
* Technical occupations defined from SAR as SOC minor codes 212, 211 and 311.82
Appendix 4: How to tell if statistics
imply discrimination
There are a number of ways of deciding whether the applications or promotion
figures give cause for concern. We first review the decision methods advocated in
the USA, of which there are three in the Basic EEO Resource Manual (Employment#p#分頁標題#e#
Policy Foundation, 1991).
USA employers are required to calculate, using one of three methods, whether there
is evidence of ‘adverse impact’ from their practices for gender groups and for race/
ethnic groups that constitute 2 per cent or more of the workforce: the four-fifths rule;
the standard deviation rule; or the 0.05 level of significance. In the USA, the
calculations do not oblige employers to use white men compared with minority ethnic
men (or white women compared with groups of minority ethnic women), but only
male applicants and employees compared with female applicants/employees, and
applicants/employees from minority ethnic groups compared with white applicants or
employees.
Four-fifths rule
The first method of assessing whether the selection process leads to adverse impact
is the simplest. An employer’s selection procedure is considered to have an adverse
英國人力資源管理課程dissertationimpact when the selection rate for any group is less than four-fifths (80 per cent) of
the rate of the group with the highest selection rate. This can be illustrated in a
simple example (see Table A4.1).
Table A4.1 Hypothetical example figures of applicants and shortlisting by gender
Selection rate
(per cent of applicants Standard
Applicants Shortlisted shortlisted ) deviation
Men 100 20 20 0.04
Women 100 10 10 0.03