Demography Special Lecture I, II

Latest update: 9 April 2014

1. Overview

Demography is one of the essential methodology to grasp the characteristics of human population survival. It can support the understanding of epidemiology and public health. This lecture basically use a textbook for basic formal demography and demographic models, followed by several examples of actual demographic researches.

Using computer practices may be helpful for understanding. Bringing your own laptop computer is recommended. Software to be used is GNU R, which is free software (

2. Plan

The lecture is done in first semester on Thursday, 13:00-14:30 at GSICS 206.

  1. Introduction
  2. Data collection: Basically 3 ways (Census, Vital Registration and Sample Survey). To collect reliable census data, properly training enumerator is essential. Identification is essential for longitudinal survey.
  3. Age and sex structure (see, Supplementary material on 25th April, R code in 20130425, with Japanese comments, R code in 20130502, with English comments, An answering code to the Exercises): In developing countries, reporting preference on 0 or 5 in last digits must be paid attention. Sex ratios are determined by the sex ratio at birth, age/sex specific mortalities, and age/sex specific migration rates. Useful measures are Whipple's index, population pyramid, and sex ratios.
  4. Period fertility (Tables in Excel format: 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7; Tables in tab-delimited text format: 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7; Important measures are age-specific fertility rate (ASFR) and total fertility rate (TFR) as the summation of ASFRs, Codes: R code in 20130509; Exercises data: 4E.1, 4E.2, 4E.3, 4E.4; Data in Excel format: Tables 4E.1-4E.4, Table 5.1-5.5; An answering R codes to Exercises)
  5. Cohort fertility (Tables in tab-delimited text: 5.1, 5.2, 5.3, 5.4, 5.5), (R code in 20130523, ), (Table 5.6 Excel Worksheet in 20130530, Table 5.6 R code), (Exercises: Excel format data, Table 5E-1 in tab-delimited text, Table 5E-2 in tab-delimited text, Table 5E-3 in tab-delimited text, Table 5E-4 in tab-delimited text, Table 5E-5 in tab-delimited text; The R code to answer the exercises of Chapter 5)
  6. Mortality and life tables: Crude Death Rate (CDR) for general mortality level (but ignoring age structure), Infant Mortality Rate (IMR) for sanitation level (Pay attention to the difference between the populations of numerator and denominator), Neonatal Mortality Rate (NMR), Early Neonatal Mortality Rate (ENMR), Stillbirth Rate (Note: Pay attention to the definition of stillbirth/spontaneous abortion, in current Japan, 22 weeks as gestational period is critical) and Perinatal Mortality Rate for medical standards and maternal health level (Note: Maternal Mortality Rate is not explained in the text.) (see, Supplementary explanation for 20130606), Explanation for "Specific Death Rates" and "Standardization": As Sweden and Kazakhstan's population and death data by age show, CDR is largely affected by age-structure and ASDRs are difficult to see. Age-standardized mortality rates are useful (R code with comments; amongst, SMR is an important measure, especially for developing countries with poor data quality). Lifetable functions are very important in demography. Needed information is basically ASDR. See. tables (Excel format, Open Document format -- or LibreOffice) and R code to calculate lifetables. The data and answers to Exercises: Excel file, R code and comments.
  7. Migration: (No exercise)
  8. Marriage and divorce: United Nation's recommendation for minimum classification of marital status: single (i.e. never married), married, widowed not remarried, divorced not remarried, remarried. Crude Marriage Rate = Marriages/Population x 1000. General Marriage Rates (females) = Marriages / (Unmarried females aged 15+) x 1000 (Note: Demominator can be All females aged 15+ when the marital status information is unavailable or unreliable), General First Marriage Rate = (Marriages to single females) / (Single females aged 15+) x 1000, Age Specific Marriage Rate (F, 20) = (Marriages to females aged 20) / (Unmarried females aged 20) x 1000, Age Specific First Marriage Rate (F, 20) = (Marriages to single females aged 20) / (Single females aged 20) x 1000. Alternative way of calculation, less popular: ASMR(F, 20) = (Marriage to females aged 20)/(All females aged 20) x 1000, ASFMR(F, 20) = (First marriages to females aged 20) / (All females aged 20) x 1000. In this manner, sum of ASFMR forms Total First Marriage Rate, which means the proportion of women who would eventually marry if the age-specific rates prevailed. Mean age at marriage and SMAM (Singulate Mean Age at Marriage) are informative. The former requires the data based on marriage resistration, the latter does not. SMAM = {(100x15 + (sum of %single for ages from 15-19 to 45-49) - (mean of %singles of aged 45-49 and aged 50-54)/2 x 50} / {100 -(mean of %singles of aged 45-49 and aged 50-54)}. CRR (Crude Remmariage Rate) = (Females remarrying) / (Divorced and widowed females) x 1000. ASRR (Age Specific Remarriage Rate) (F, 30) = (Females aged 30 remarrying) / (Widowed and divorced females aged 30) x 1000. Please see R code for text and R code for exercises. The exercises for this chapter are difficult.
  9. Reproducibility (25th July): Reproducipility means how much the generation is reproducing itself. New period measures of fertilities considering the effect of sex and mortality. GRR (Gross Reproduction Rate) sums up the age-specific fertility rates for only females' births (similar to TFR, but the numerator is limited for females' births; it's approximately same as TFR multiplied by the proportion of females at birth). NRR (Net Reproduction Rate; denoted by R0, pronounced as [ɑr nɔːt]) is a GRR adjusted for mortality: by multiplying each female ASFR by the probability of survival from birth to that age before summing up. The probability of survival from birth to age x is lx of lifetables. 5Lx, the person-years survived from exact age x to x+5 (in other words, the stationary population for each age-class associated with the life table), will be (l[x]+l[x+5])/2*5. Then female births to women in stationary population can be calculated by the product as (female ASFR)x(5Lx). Sum of these gives NRR. The interpretation of NRR is the average number of daughters a woman would have during her reproductive years assuming fertility and mortality schedules to be fixed. If NRR<1, the population in the synthetic cohort will decrease after considerable years. To be noted, the decrease does not necessarily happen in any actual cohort. Postwar Europe experienced the rapid rise in period fertility, when the NRR caused misleading. That's why TFR is more popularly used than NRR. Alternative calculation of NRR is, GRR multiplied by the probability of surviving to the mean age of the age-specific fertility distribution (lx at mbar). It's a good approximation. About the timing of fertility, the mean age of childbearing (μ) has become commonly used after mbar. μ is also the weighted mean of mid-points of the age groups, as same as mbar, where the weights are not females ASFR but the births to the stationary population. μ is generally slightly less than mbar, because μ considers the effect of mortality. Median ages of the age-specific fertility distribution (m) and the mean ages of mothers giving births (Mbar) are sometimes used. When considering stable population instead of stationary population, mean length of generation can be considered (see, Chapter 11). R code for text, R code with comments for exercise.

* Examples of demographic analysis in Senegal or other countries (lecture given by Dr. Reiko Hayashi, on 18th July)

Chapters below, included in "Part II: Demographic models". In this lecture, we had no time to learn these in 2013.

3. Evaluation

Based on presentation, discussion, and report.

4. Textbook

Newell C (1988) Methods and models in demography. The Guilford Press

(Reference) Preston SH, Heuveline P, Guillot M (2001) Demography: Measuring and Modeling Population Processes. Blackwekk Publishing.

5. Office hour

For the students of the Graduate School of Health Sciences, Tuesday, 18:00-18:30, at Myodani campus E707. For the students of GSICS, Thusday, 16:40-18:00 at Frontier Building Room 717. Taking appointment is recommended.

6. Message to the students

Done in English.

Notice to cite or link here | [TOP PAGE]