Docstoc

Paul_Norman_presntation

Document Sample
Paul_Norman_presntation Powered By Docstoc
					       The Dynamics of Populations Large & Small:
                   Theory & Applications


   UK subnational variations in fertility &
     infant mortality: 1981 to 2006

                       Paul Norman
                    School of Geography
                     University of Leeds

Acknowledgements
• Phil Rees, Pia Wohland, Pete Boden, Nik Lomax
• Paul Boyle, Ludi Simpson, John Stillwell
• ESRC, ONS, GROS, NISRA, OS, UKBORDERS, CDU
Prequel ...
University of Liverpool, September 1977
Joint conference
• British Society of Population Studies
• Population Study Group of the Institute of
British Geographers

Regional variations in fertility & child mortality
during the demographic transition in England &
Wales (study period c. 1860s – 1920s)

William Brass (1921–1999)

“There is no geography to fertility”

(Rees, pers. comm.)
Prequel ...
Regional Demographic Development (1979)
Edited by John Hobcraft and Philip Rees

Regional variations in fertility & child mortality
during the demographic transition in England &
Wales (Ch. 3)

William Brass & Mohammad Kabir

“The single most notable finding is the
astonishing consistency of the falls in
fertility by regions and the almost total
preservation of the relative differentials of
1876, fifty years later” (p. 87)
Demographic Transition
Model which describes population change over time

                                                                   Thompson (1929)
                       Birth rate
                                                                   Landry (1934)
                                    Population
                                                                   Notestein (1945)
                Death rate




                                     Mature-        Post-
   Pre-modern     Urbanising
                                    industrial    industrial




Mortality decline: consensus                     Fertility decline: reasons speculative
• Improved hygiene & living                      • Urbanisation increases costs of children,
standards                                        rationality
• Fewer deaths from violence & war               • 19th Century, unlikely to be contraception
• Reduction in infectious disease                & change in family values
• Lower infant mortality                         • Fertility decline follows mortality decline …
Fertility & infant mortality, 1850–1930
            England & Wales, 1850-1930




                               Brass study




William Brass
• Does lower ‘child mortality’ precede lower fertility?
• ‘Regions’ and counties in England & Wales, c. 1860s–1920s
Fertility & child mortality, 1860s–1920s
Fertility decline influenced by infant mortality decline?

   Regions                                        • Highest correlations should be
                                                  in the next two to three years
                                                  • Maximum c. 0 displacement
                        Fertility change af ter
                        mortality change          • “Time-localised correlations
                                                  of trends in areas are entirely
                                                  consistent with the view that
                                                  causative factors were largely
                                                  the same for fertility as for
   Counties                                       child mortality” (p. 86)

                                                  But …
                                                  • Falls in fertility by region
                        Fertility change af ter
                        mortality change          almost uniform
                                                  • Declines in child mortality vary
                                                  geographically
Second Demographic Transition?
Rees (1997)

First demographic transition
• State of high birth & death rates …
    … to low rates
• Zero or slow population increase …
    … to rapid growth

Second demographic transition
• Low birth rates lower than the low death rates
• Population begins to decrease …
    Mitigated / postponed by
    • Immigration to destination country
    • Demographic momentum & increased life expectancy
Fertility & infant mortality, 1981–2006
Subnational fertility generally declining, largely due to ‘postponement’
(Tromans et al., 2008)
Recent concerns over low fertility, especially in Scotland (Boyle, 2003;
Boyle et al., 2007 )
Infant mortality declining, but not even geographically (Norman et al.,
2008)
               Total Fertility Rates             Infant Mortality Rates




            Northern Ireland: ASFRs                Scotland: ASFRs
 Fertility & infant mortality, 1981–2006
TFRs




       1981   1986   1991   1996   2001    2006


IMRs
Fertility & infant mortality, 1981–2006
1981 to 2006 by UK local authorities by ONS Supergroups
           Total Fertility Rates             Infant Mortality Rates




           TFR relative to 1981               IMR relative to 1981
Fertility change
Subnational fertility generally declining, largely due to ‘postponement’
(Population Trends 133, Tromans et al., 2008)
Fertility & infant mortality, 1981–2006
Relationship between (change in) infant mortality and fertility?

    Local                                       • Infant mortality & fertility
    authorities




    Local                                       • Change in infant mortality &
    authorities       Fertility change af ter   change in fertility
                      mortality change




                                                • Suggests different ‘causative’
                                                factors
Fertility & infant mortality, 1981–2006
Factors relating to fertility & infant mortality?

    • Social Class                    • Ethnicity
    • Educational achievement         • Students
    • Unemployment                    • Armed forces
    • Tenure                          • Country
    • Deprivation

Fertility & infant mortality can be modelled in 1991 & 2001
    • Consistent geography & variables
Area variation in fertility
1991 TFR        Adj R square = 0.58                 ***     B      B (std)   Sig.
                Constant                                   +1.20
Adj R sq 0.58                                       ***
                Townsend index                             +0.01   +0.32     ***
                                                     *
                log % Students                             -0.07   -0.36     ***
                                                   ***
                log % Pakistani or Bangladeshi ethnicity   +0.04   +0.48     ***
                log % Chinese ethnicity             ***    -0.09   -0.34     ***
                log Persons per hectare             ***    -0.01   -0.22     ***
                Wales                                      +0.02   +0.08      *
                                                    ***
                Scotland                                   -0.07   -0.34     ***
                                                    ***
                Northern Ireland                           +0.11   +0.32     ***



                                                    ***
2001 TFR        Adj R square = 0.57                         B      B (std)   Sig.
                                                    ***
                Constant                                   +1.23
Adj R sq 0.57   Townsend index                      **     +0.01   +0.26     ***
                log % Students                      ***    -0.11   -0.55     ***
                log % Pakistani or Bangladeshi ethnicity   +0.05   +0.64     ***
                                                   ***
                log % Chinese ethnicity                    -0.04   -0.17     ***
                                                   ***
                log Persons per hectare                    -0.01   -0.23     ***
                Wales                               ***    +0.03   +0.09      **
                Scotland                            ***    -0.05   -0.22     ***
                Northern Ireland                           +0.09   +0.22     ***
Area variation in infant mortality
1991 IMR
                                   ***
Adj R sq 0.33
                 Adj R square = 0.33                          B      B (std)   Sig.
                 Constant                                   +1.26
                 Townsend index                             +0.01    +0.22      ***
                                   ***
                 log % Low Social Class                     +0.23    +0.23      ***
                 log % Pakistani or Bangladeshi ethnicity   +0.12    +0.32      ***


                                   ***




2001 IMR                           ***

Adj R sq 0.36   Adj R square = 0.36                          B      B (std)    Sig.
                Constant                                    +0.74
                Townsend index     ***                      +0.02   +0.26      ***
                log % Low Social Class                      +0.34   +0.25      ***
                log % Pakistani or Bangladeshi ethnicity    +0.11   +0.28      ***


                                   ***
Unexplained variation in TFRs
       1991                                                2001

TFRs    Residuals                                   TFRs     Residuals




                             1991 cf 2001
                             TFRs = 0.77 ***
                             Residuals = 0.58 ***




                    Rates lower than predicted
Unexplained variation in IMRs
       1991                                                2001

IMRs    Residuals                                   IMRs     Residuals




                              1991 cf 2001
                              IMRs = 0.32 ***
                              Residuals = 0.11 **




                    Rates lower than predicted
1860s-1920s cf. 1981-2006 …
1860s–1920s, Brass …
• Falls in fertility by region, preservation of relative differentials
• Declines in child mortality vary geographically
• Time-localised correlations … causative factors largely the same for
fertility as for child mortality

1981–2006
• Variations over time in fertility very similar across LAs, differentials largely
preserved. Declines in infant mortality some variation
• Weak or no correlations between TFR & IMR suggest different ‘causative’
factors
    • For 1991 & 2001 some influential variables in common: deprivation &
    Pakistani & Bangladeshi ethnicity. London an ‘outlier’?
• Why the rise in fertility between 2001 & 2006? Will this continue?
    • Economic boom? Demographic momentum? Immigration?

• Study ‘health warnings’: causality & ecological fallacy, global OLS
regression, local authority geography, etc.

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:4
posted:6/24/2011
language:English
pages:18