Race and Population a Statistical Demographic Approach

1
Race and Population a Statistical Demographic
Approach
Preliminary findings in investigating and
estimating fertility rates by ethnic group within
small areas

• Lee Williamson
• CCSR
• Cathie Marsh Centre for Census Survey Research
• The University of Manchester
• Lee.williamson_at_stud.man.ac.uk

2
Background

• The PhD is CASE partnered with Bradford council
• Most of the research will be primarily carried
  out using data provided by Bradford Council
• Core problem of creating demographic rates and
  population projections where there is very little
  data available
• (small areas, ethnic groups or both)
• This presentation will focus only on creating
  fertility rates by ethnic group within small areas

Introduction Data Grouping
Method Results
3
Research overview

• Statistical methods will be used to extend robust
  estimates of demographic rates to small areas and
  ethnic groups.
• Different strategies of both methodological
  approach and data sources will be implemented.
• The impact of each will be assessed by
  implementing the estimates in projection software
  Popgroup provided by Bradford Council.

Introduction Data Grouping
Method Results
4
Research outcomes

• The outcomes are to provide
• Demographic rates and population projections
  estimated for wards of Bradford, ethnic groups in
  wards of Bradford, and ethnic groups in Great
  Britain.
• Robust and documented estimation strategies for
  sub-national demographic rates for ethnic groups,
  applicable to any area in Britain.

Introduction Data Grouping
Method Results
5
Bradford Metropolitan District

• Bradford is a multicultural district
• Made up of 30 wards
• Ward sizes range
• from 13,000 to 23,000
• Total population 480,000
• (1991)

Introduction Data Grouping
Method Results
6
Data sources for 1991 fertility rates

• Age-specific fertility rate
• fx births in year t to women aged x last
  birthday at the time of birth
• mid-year population of women aged x last
  birthday
• Denominator population is 1991 population
  estimate by ethnic group West Yorkshire Council's
  work commissioned from GMAP Ltd (Leeds
  Universitys consultancy company)

Introduction Data Grouping
Method Results
7
Background to births data

• The numerator of the births originates from the
  Bradford Birth Statistics Database which was
  originally is compiled and provided by Bradford
  Community Health Trust and the Airedale NHS
  Trust.
• Included in the data set were the variables of
  interest
• Date of birth of baby
• Date of birth of mother
• Ethnic group of the baby
• Ward

Introduction Data Grouping
Method Results
8
Background to births data

• The data set did require some cleaning
• Initially the ethnic group categories provided by
  in the documentation were not the same as those
  that were to be used in the denominator to create
  the fertility rate.
• The 6 ethnic groups that the rates are to be
  created for are
• White Black (Caribbean, African and Black
  Other)
• Indian Pakistani
• Bangladeshi Other
• Ideally this would result in the creation of 180
  sets of ASFRs that is 30 wards by the 6 ethnic
  groups.

Introduction Data Grouping
Method Results
9
Implication for 5 years 1989-1993

• Taking 5 years births records from 1989-1993 and
  averaging them to be centred on 1991
• Aim to reduce fluctuations in comparison with
  using 1 years worth of birth records
• 37,557 births over the 5 years
• Not all records had valid ward codes sensible
  age of mother aged between 12-50
• Resulting in 36,173 valid births records over the
  5 years which will be averaged and centred on
  1991 to create the rate

Introduction Data Grouping
Method Results
10
Operationalising the birth data

• Data for small areas and ethnic groups are often
  aggregated up to
• 5 years of ages together due to small numbers
  (15-19, 20-24)
• Why aggregate data when you have it by single
  year of age (SYOA)

Introduction Data Grouping
Method Results
11
ASFR curves

• Fertility schedules for the example wards

Introduction Data Grouping
Method Results
12
Clustering wards based on 1991 census variables

• population aged under16
• non-White in ward
• of those reporting limiting long-term illness
• 16 plus people who are economically inactive
• of households renting from LAs
• of households with no car
• of households with no facilities
• of households with a density of more than 1
  persons per room

Introduction Data Grouping
Method Results
13
Classifications from ONS deprivation indices
Introduction Data Grouping
Method Results
14
Mapping of Bradford ward TFRs
Introduction Data Grouping
Method Results
15
Smoothing the fertility rates

• 20 sets of fertility rates were smoothed using
  the Hadwiger curve
• Some sets of ASFRs are still very jagged
• White Black Indian
• Suburbia Suburban more established areas

Introduction Data Grouping
Method Results
16
Smoothed fertility curves

• Ethnic group White Ethnic group Black

Introduction Data Grouping
Method Results
17
Smoothed fertility curves

• Ethnic group Indian Ethnic group Pakistani

Introduction Data Grouping
Method Results
18
Smoothed fertility curves

• Ethnic group Bangladeshi Ethnic group Other

Introduction Data Grouping
Method Results
19
Testing these preliminary results

• Projections were created in Popgroup using 3
  rates
• GAD England
• Bradford district ASFR
• Hadwiger smoothed group ASFR
• Projected births compared with actual births over
  a 9 year period

Introduction Data Grouping
Method Results
20
Conclusions of preliminary investigations

• The influence on Bradford wards fertility
  behaviour is complex,
• no simple relationship between the level and
  ethnic group.
• Characteristics of the ward also play a key role
  in the fertility level.
• The 20 ward groupings are far from ideal.
  However, given the small numbers involved it
  provides the most logical option available.
• Next step is to test the robustness of the wards
  inclusion in each group, plus to replace
  group rates where there are sufficient numbers
  of births for the ward.

Introduction Data Grouping
Method Results