Analysis of socio-economic factors influencing fertility in South by gyvwpsjkko

VIEWS: 10 PAGES: 23

									Analysis of socio-economic factors influencing
            fertility in South Africa

                              Rosinah Senona
                          Statistics South Africa
              Institute of Statistics and Applied Economics,
                           Makerere University

 A paper prepared for presentation at the 1st Africa Conference of Young
                            Statisticians 2008
                 Pretoria, South Africa, 1 – 3 July 2008


                                                                       1
Background
• High population growth rates has been acknowledged by
  nations as resulting from high fertility levels and one of
  the impediments to rapid social and economic
  development
• Declining population growth rate as resulting from low
  fertility due to modernization and developments
• Less developed countries are attempting to reduce the rate
  at which their populations are growing, while developed
  countries are concerned with falling fertility rate
• It is evident that the design of effective and efficient
  policies requires knowledge of the determinants fertility.


                                                           2
  Research context
• Previous research on fertility in South Africa has
  primarily focused on fertility transition, fertility
  levels and trends and policy implications.
• Since the end of apartheid era, most researchers
  were interested in unraveling the age and parity
  structure of the country’s fertility transition.
• Other studies attempted to fill gaps that existed
  in South Africa’s demographic history by
  focusing on contemporary features of South
  African fertility.
                                                     3
Research context-cont’d
• Although previous studies on fertility shed light on fertility
  status on all racial groups and brought attention to the spatial-
  temporal aspects of population redistribution, a better
  understanding of population dynamics can be gained if the
  relationship between fertility and socio-economic relating to it
  are examined.
• To our knowledge, few studies has empirically examined the
  relationship between indirect socio-economic factors and
  fertility in South Africa using the first Demographic Health
  Survey data conducted in 1998.


                                                                      4
Research context-cont’d
• Now with the availability of DHS data,
  this study contributes to the discussion on
  fertility by exploring the impact of socio-
  economic factors on fertility in South
  Africa.




                                                5
Conceptual framework
  Indirect determinant           Direct determinants               Fertility
  Socio-economic                  Proximate
  factors                         determinants
  Employment                      Contraceptive use
  Education                       Marital status
  Income                          Age at first                   Children ever born
  Place of residence              marriage
  Race                            Desire for children
• The analytical framework of the proximate determinants by Bongaart (1978, 1985)
   provides us with a deeper understanding of how socio-economic, biological and cultural
   factors affects fertility.
• In this study, we explore the empirical evidence of these factors on fertility and
   attempt an explanation for the observed pattern.


                                                                                    6
 Methodology
• The study is based on secondary data from the
  Demographic Household Survey, 1998.
• The DHS is a nationally representative, stratified, self-
  weighting proportional to size sample survey of women
  aged 15 to 49 years.
• The sample size of about 11735 was used but the analysis
  in some case was limited to women with income which
  yielded a sample size of 4023 eligible women.
• Statistical Package for Social Scientist was used in fitting
  the model
                                                           7
Regression Analysis
• Because children ever born is a non-negative count
  variable, count data models are the natural choice for
  the regression.
• Some considerations on regression models with count
  data were done in choosing which model is
  appropriate for the dataset.
• We considered generalized Poisson regression model
  as it has statistical advantage over both standard
  Poisson and negative binomial.
• The model is approximated by maximum likelihood
  method.

                                                       8
Regression Analysis-cont’d
• The key concern with this model is the
  assumption of constant variance
  (Homoscedasticity) which was violated.
• In our case, we used robust standard errors which
  adjust for hetescedasticity in the model.
• We transform coefficients by exp{B} to get
  meaningful interpretation.



                                                9
Findings

Parameter     B        Std      Hypothesis Test
estimates              Errors
                                Wald Chi-   Df    Sig
                                Square
(Intercept)   1.201    0.1033   135.116     1     0.000

Employment    0.252    0.0423   35.468      1     0.000

Income        -0.116   0.0766   2.285       1     0.131

Place of      0.131    0.266    24.273      1     0.000
residence
Level of      -0.278   0.0167   277.430     1     0.000
education
Race          -0.152   0.0168   82.338      1     0.000


                                                          10
Findings-cont’d


Parameter     B        Std      Hypothesis Test
estimates              Errors   Wald Chi-         Df   Sig
                                Square

(Intercept)   1.098    0.0428   657.035           1    0.000
Employment    0.323    0.0179   327.919           1    0.000
Pace of       0.133    0.0190   48.880            1    0.000
residence
Level of      -0.471   0.0118   1607.787          1    0.000
education
Race          -0.005   0.0126   0.187             1    0.666


                                                               11
Findings-cont’d

Parameter         B        Std      Hypothesis Test
estimates                  Errors
                                    Wald Chi-         Df   Sig
                                    Square

(Intercept)       1.094    0.0384   812.203           1    0.000

Employment        0.320    0.0176   328.903           1    0.000

Place of residence 0.133   0.0186   50.892            1    0.000


Level of          -0.471   0.0116   1654.827          1    0.000
education




                                                                   12
Findings-cont’d

• Comparing the fitted model with observed data

Goodness of fit
Likelihood Ratio Chi-   Degree of freedom   Significance
Square.
868.599                 5                   0.000




                                                           13
Interpretation of the analysis

Parameter estimates   Beta     Expected log count



(Intercept)           1.094

Employment            0.320    0.377

Place of residence    0.133    0.142

Education             -0.471   -0.38




                                               14
Estimated marginal means
Parameter estimates   B      Std Errors

Rural                 1.71   0.021

Urban                 1.95   0.027




                                          15
Conclusion
• This paper sought to study the impact of selected
  socio-economic factors on fertility and provide an
  explanation for the observed pattern.

     The results shows that educational level affects
     fertility negatively suggesting that women prefer to
     further their education rather than increasing number
     of children.
     Employment and place of residence affect inversely
     fertility choice of individual women.
     Finally income and women’s race were found not
     affect women’s fertility decision.
                                                             16
Reasons for low fertility in South Africa

Very low fertility is therefore caused by the combination of several factors:
    High rates of unemployment, or unstable employment
•   Instability and insecurity
•   low standard of living
•   inadequate housing
•   lack of relevant investment in social programs
•   and difficulty in balancing career and parental
    responsibilities

                                                                           17
 Population dynamics

• An increase in population growth can be considered a
  major factor in promoting economic growth because it
  results into increased labour force.
• Increased labour force could lead to an increase in output
  produced.
• An increase in a country’s population can further
  contribute to increasing production through expanding of
  the potential size of domestic markets.
• As a consequence, domestic producers increase their
  output of goods and services in response to increased
  demand.

                                                        18
Population dynamics-cont’d
• Decline in fertility levels leads to a decrease in
  population growth.
• This significant population feature raises concern on the
  changes in the structure of the labour market and
  national economy.
• If these features are not addressed, they may result in
  the below replacement trap which were found to
  generate negative population momentum, that is a new
  force for population shrink-age over the coming decades
  due to the fact that below-replacement fertility will soon
  results in declining numbers of potential parents

                                                        19
Policy implications
• Government investment in social services
  would be a more effective way to maintain
  fertility rates.
• Democratic government adopted population
  policies which concentrated on improving and
  developing welfare of women in general.
• This paper sought to direct these policies on
  specific areas to focus on, in order to achieve
  the expected goals.
                                                20
Policy implications
To improve the welfare of women by:
Creating job opportunities
• Improving social services both in rural and urban
  areas and more emphasis should be in closing
  the gap in social services that exist between
  rural and urban areas.
• Effective and efficient programs to educate and
  empower women in all areas.
• Effective monitoring and evaluation systems in
  place to ensure both quantity and quality
  outcomes.
                                                  21
EMPOWER WOMEN, SAVE THE NATION




                                 22
Thank you




            23

								
To top