Mortality in Russia: Evidence from Micro Data - PowerPoint PPT Presentation

1 / 15
About This Presentation
Title:

Mortality in Russia: Evidence from Micro Data

Description:

Family in the changing world, November 28-29, Moscow. Mortality in Russia: Evidence from Micro Data. Irina Denisova ... Center for Economic and Financial ... – PowerPoint PPT presentation

Number of Views:18
Avg rating:3.0/5.0
Slides: 16
Provided by: bfmi
Category:

less

Transcript and Presenter's Notes

Title: Mortality in Russia: Evidence from Micro Data


1
Mortality in Russia Evidence from Micro Data
  • Irina Denisova
  • Center for Economic and Financial Research at the
    New Economic School, Moscow

2
Motivation
  • Relatively high and only slightly declining
    within the latest decade mortality rates
  • Mortality crisis in 1992-1994
  • High mortality rates among working age population
  • Factors behind?
  • Environmental
  • Behavioral

3
(No Transcript)
4
Literature
  • Two branches of literature
  • Aggregate data analysis (mortality registration
    data, census data)
  • Micro data analysis (follow-up surveys, large
    panel surveys)
  • Russia
  • Aggregate data
  • Basic patterns of mortality crisis
  • Male Cohort within US Lipid Research Clinics
    Program

5
Research questions
  • What are the determinants of mortality rates
    (hazard and survival rates)?
  • What is the role of different groups of factors?
  • hazardous behavior (negative investment into HC)
  • income
  • relative income (relative deprivation and
    unfairness)
  • stress factors
  • poor social capital, incl. family characteristics
  • job-related factors

6
Data
  • Russian Longitudinal Monitoring Survey
  • Waves 5 14 (1994 to 2005)
  • Nationally representative, about 5,000 households
    and 10,000 individuals in each round
  • Panel structure though attrition is a serious
    issue
  • If a household member is missing, reasons for
    being not a hh member
  • Moved out, another address
  • Separate hh now, same address
  • Death (cause of death since 2001)
  • Other reasons

7
(No Transcript)
8
Methodology
  • Survival analysis
  • Eliminates estimation bias due to non-normality
    of time to event and right-censoring
  • Hazards are estimated for uncensored and
    survivals for censored
  • Allows using an unbalanced panel
  • Proportional hazard model
  • Parametric PH model, Gompertz specification
  • ?0(t) exp(?t) exp(ß0)
  • Cox PH model specification (non-parametric
    baseline)

9
Methodology
  • Explanatory variables (X) include
  • Gender
  • Education and qualification
  • Settlement type (urban vs rural)
  • Smoking, alcohol consumption vs physical
    exercises (hazardous vs healthy behavior)
  • Characteristics of diet
  • Income decile
  • Relative income position (relative deprivation
    and unfairness)
  • Relative respect and power positions
  • Stress factors (years in unemployment, concern
    about getting necessities, satisfaction with
    life)
  • Job characteristics (hazardous working
    conditions)
  • Social capital, incl. family characteristics
    (family size, marital status, children)
  • Two samples adults and adults 20-60 years

10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
Baseline survival function
15
Results
  • Subjective perception of relative deprivation is
    an important determinant
  • Higher income deciles are correlated with longer
    life (could be health effect good health brings
    higher income and longer life)
  • Smoking survives all specifications as a
    detrimental factor
  • Heavy alcohol consumption also matters
  • Better education is beneficial
  • Role family social capital is not very pronounced
    and operates in opposite directions
Write a Comment
User Comments (0)
About PowerShow.com