Population Structure and Life Expectancy: Community Area Dynamics in Leeds, 19912001

1
Population Structure and Life Expectancy
Community Area Dynamics in Leeds, 1991-2001

• Phil Rees, John Stillwell and Amy Tyler-Jones
• School of Geography, University of Leeds
• Paper presented at the Second International
  Conference on Population Geographies, RGS (with
  IBG) Population Geography Research Group, 11-14
  August 2004, University of St. Andrews, Scotland

2
Context and Aims

• Twenty-First Century Leeds Geographies of a
  Regional City, edited by Rachael Unsworth and
  John Stillwell, published by the School of
  Geography in celebration of the University of
  Leeds Centenary (1904-2004)
• Aims of paper (part of book)
• To characterise Leeds neighbourhoods in terms of
  their age and household structures
• To characterise one important element that is
  changing that structure mortality, by estimating
  how life chances vary by neighbourhood in Leeds

3
Population pyramids Leeds vs Eng Wales, 1991
2001

• 1991
• High early century births
• Late 40s bulge
• 1955-1971 baby boom
• Post 1971 baby bust
• 1986-91 revival
• Leeds student/yuppie surplus
• 2001
• Leeds younger than EW
• Pyramid evolving to urn
• Yuppies stay on
• Students have grown

4
The changing sex ratio (males per 100
females)Leeds and England Wales, 1991 2001

• Natural male surplus
• Missing men (emigrated without their women)
• Better female survival chances
• But males are catching up

5
What happens at the neighbourhood scale?Where
are the Leeds communities?

• City council survey leads to 100 communities
• Leeds team (Stillwell, Tyler-Jones, Shepherd)
  match neighbourhoods to 2001 Census Output Areas
• Six extra Communities added to give detail to
  rural periphery
• 106 Community Areas on the map

6
The geography of household types (1) single
person households

• Single person households are concentrated in the
  inner city and inner suburbs

7
The geography of household types (2) family
households, all pensioner

• All pensioner households are concentrated in the
  outer suburbs to the north of the city

8
The geography of household types (3) family
households, married couples

• Married couple household are concentrated in the
  outer suburbs with recent housing developments

9
The geography of household types (4) family
households, cohabiting couples

• Cohabiting couple family households are
  concentrated in the southern and western suburbs
  (where there is affordable housing)

10
The geography of household types (5) family
households, lone parent

• Lone parent households are concentrated in inner
  city, low cost housing neighbourhoods

11
The geography of household types (6) Other
households

• Other households (multi-member,unrelated) are
  concentrated in the inner suburbs to the north of
  the city centre

12
The Geography of Age GroupsChildren (0-15 years
old)

• Multi-ethnic neighbourhoods
• Social housing estates
• Newer suburban housing
• The student quarter
• 1991 Census
• 2001 Census

13
The Geography of Age Groups16-64 years of age

• The student quarter
• The outer suburbs
• 1991 Census
• 2001 Census

14
Students are the new Blacks (The Economist,
2004)

• University centred residential distribution
• Differences between locations of UG year ones and
  others and PGs
• Communal Housing linked to transport axis
• Town-gown conflicts and protests about Multiple
  Occupancy Housing
• New policy for University build southward and
  regeneration linked

15
The Geography of Age Groups65 years and older

• Concentration outside the inner city and social
  housing estates
• Shift outwards over the decade
• Combination of ageing in place and some
  out-migration
• (a) 1991 Census
• (b) 2001 Census

16
Institutional facilities for the most infirm
old and the convicted

• 1.3 of the 2001 population lives in communal
  establishments
• Rees senior
• There is both private and social provision
• Includes nursing homes, old persons homes,
  childrens homes, younger offenders institutions
  adult prisons

Communal Establishments
17
Changing dependency ratios 1991-2001

• The dependency ratio is defined here as 100 times
  (population aged 0-15 population aged
  65)/(population aged 16-64)
• There are both increases and decreases
  increasing dependency ratios are not universal
• Ageing in the 1991-2001 in the UK was subdued
  (1920-40 cohorts were smaller than 1900-1920
  cohorts)
• Note student quarter becomes more adult

18
A demographic classification of community areas
19
Population pyramids for 10 clusters

• Variables percents in 5 year ages
• Method k-means clustering (SPSS)
• Ten cluster solution chosen
• Outliers Extreme students, City Centre (hostels,
  new apartment complexes)

20
Distribution of community area cluster members

• The clusters are distributed as we might expect
  given previous age group maps
• On balance the household classification gives a
  better picture of way different ages live
  together and live apart

21
Life expectancies at neighbourhood scale

• Life expectancies are the best measures of
  mortality experienced by a population
• There are difficulties in estimating life
  expectancies for small populations, for which we
  propose solutions
• Assuming the solutions are robust, we try to
  answer the following questions
• What are the life chances of the people of Leeds
  around 2001?
• How do these differ between men and women?
• How do they vary across the city?
• How do they compare with the national measures?
• Have life chances diverged or converged across
  neighbourhoods in the last decade?

22
Method for estimating life expectancies for wards

• Compute three year average mortality rates for
  England and Wales, 1990-1992 2000-2002, using
  1991 and 2001 mid-year estimate populations
• Use electoral ward deaths information for
  1990-1992 and 2000-2002 for ages 0, 1-4, 5-14,
  25-34, 45-54, 55-64, 65-74, 75-84 and 85
• Multiply the England Wales quinary mortality
  rates by the ratio of the observed deaths in the
  larger age group they are members of divided by
  the number of deaths in the larger age groups
  expected from the sum of E W death rates times
  the ward quinary age population
• This borrows information from a large population
  but uses critical information from the small area
  population

23
A life table for males, Cookridge ward, Leeds,
2000-02
24
The spatial pattern of life expectancies for
males, (a) 1990-92 and (b) 2000-02

• Northern and Eastern suburbs favoured
• Spatial pattern very stable over 10 years
• e0 improves 2.68 years for men, 2.50 for women
  over 10 years
• For England Wales, over the 10 years mens e0
  rose by 2.57 years and womens by 1.67 years

Where I board the bus in the morning (Cookridge
ward), male life expectancy is 79 years
(2000-02) Where I alight from the bus (University
ward), men can expect to live only 71 years.
25
Summary indicators of life expectancies, Leeds,
1990-92 and 2000-02

• Increases in life expectancies in all wards
  except one, City Holbeck fro me (problems of
  the homeless)
• But inequality has increased on all measures, SD,
  IQR and range
• Top third of wards saw 3.7 years improvement for
  men, 3.4 for women
• Bottom third saw only 1.5 years improvement for
  men and 2.3 for women

26
Estimation of life expectancies for community
areas

• Method regress life expectancy against Townsend
  deprivation score for 33 wards
• Use equation with Townsend scores for 106
  community areas (CAs) to predict life
  expectancies for community areas
• Adjust the predictions so the weighted sum of CA
  life expectancies is equal to the ward estimate
  for life expectancy
• Allows CA life expectancies to vary around their
  ward means
• Assumes relationship at ward scale holds at CA
  scale

27
Neighbourhood life expectancies, 2000-02 and
change 1990-92 to 2000-02

• (a) men, (b) women
• Poverty and inner city living seriously affect
  your health
• Poorer communities are gaining less than richer
• Inequality is greater for men than women and it
  is increasing more for men

28
Conclusions Demographic Structures

• The age-sex structure of the Leeds population was
  very stable between 1991 and 2001, resembling
  that for the country as whole (except for a
  greater share in student ages)
• Unusually, the population did not age
  significantly and there were decreases in the
  population shares which were female and elderly
• There was a big increase in the student age
  population and increasing intensity of student
  occupation of inner North West Leeds

29
Conclusions spatial distributions

• The citys communities are moderately
  differentiated in terms of household and age
  structure. Each household type is concentrated in
  a different section of the city, though all types
  and all ages are found in all communities.
• Dependency ratios have changed consequent on the
  changes in age groups both upwards and downwards.
  The largest shifts upwards have been in selected
  outer suburbs (ageing in place and out-migration
  of richer retirees). The largest shifts downwards
  have been in the student quarter and adjacent
  neighbourhoods.

30
Conclusions life chances

• Life expectancies in Leeds have improved faster
  than in England and Wales as a whole,
  particularly for women
• However, men have improved their life chances
  more than women in all neighbourhoods
• Areas at all socio-economic levels have seen
  improvements but more at the top than at the
  bottom, indicating widening inequalities
• One exception is the city centre ward, where
  homeless men face poorer life chances at the
  start of the 21st century than they did ten years
  before