Workshop on Demographic Analysis and Evaluation Fertility Indirect

Workshop on Demographic Analysis and Evaluation

Fertility: Indirect Estimation Based on Age Structure. Rele’s Method

Introduction to Indirect Estimation of Fertility No two countries have the same data for demographic analysis. Some countries have a wealth of data from vital registration systems, ongoing survey systems, and censuses. But even for these countries, data quality from survey to survey, and from census to census, may vary. Other countries collect less data and the data collected may be more likely to be incomplete or deficient in some way. The toolkit of techniques specifically designed for use where vital registration system coverage of births is incomplete -- “indirect estimation methods” – includes several based on population composition. 3

Rele’s Technique In this part of the workshop we will cover: Indirect estimation of fertility using child-woman ratios and estimated mortality -- Rele’s technique. This method is described in Population Analysis with Microcomputers, chapter 4 (pp. 204 -205 and appendix IV-5). 4

Rele’s Technique The Rele (1967) technique is based on the observation that the ratio of children to women is related to a certain fertility level. The Rele technique estimates a total fertility rate for one or two 5‑year periods prior to the census date based on data on children and women by age, and life expectancy at birth for both sexes. The data on age are used to calculate various child‑woman ratios. 5

Rele’s Technique The important aspect of these ratios is the age of the children. Ratios based on children ages 0 to 4 provide estimates of fertility for the 5 -year period prior to the information date, while ratios based on children ages 5 to 9 provide estimates of fertility for a period 5 to 10 years prior to the information date. The estimates are based on the following four ratios: (1) The ratio of children age 0 to 4 to women age 15 to 44. (2) The ratio of children age 0 to 4 to women age 15 to 49. (3) The ratio of children age 5 to 9 to women age 20 to 49. (4) The ratio of children age 5 to 9 to women age 20 to 54. 6

Rele’s Technique Rele observed that in stable populations, at a given level of mortality, there is an almost linear relationship between the child/woman ratio and the fertility level. Starting from this observation, he fitted regression equations linking the gross reproduction rate with such ratios for different levels of life expectancy at birth taken from Coale-Demeny stable populations. Hence, given a life expectancy at birth and child/woman ratios, an estimate of the gross reproduction rate can be obtained by using fitted coefficients. And a predicted TFR can be obtained if the sex ratio at birth is known or can be assumed. 7

Rele’s Technique Rele’s regression coefficients are shown below. 8

Rele’s Technique Data required: The child population (both sexes combined) ages 0 to 4 and 5 to 9. The female population ages 15 to 44 years, 15 to 49, 20 to 49, and 20 to 54. The life expectancy at birth for both sexes combined. 9

Rele’s Technique Rele’s technique is illustrated here using the Census Bureau’s PASEX spreadsheet RELEFERT. xls, and 1970 census data for a mythical country, Orientia Spreadsheet: RELEFERT. xls 10

Rele’s Technique Output of RELEFERT. xls 11

Rele’s Technique The analyst then has the following tasks: 1. Review results for each reference date and associated age group, as these connect to the population input. 2. Consider the strengths and weaknesses of the population input data, and how they impacted each result. Your analyses on age and sex structure will help in this assessment! 3. Identify the most accurate measure(s) accordingly. 12

Rele’s Technique Assumptions: The population is stable or quasi-stable. There is no migration. Completeness of enumeration is the same in all the required age groups. Age reporting of children is correct within each 5 -year age group. Age misreporting of women within the groups of childbearing ages does not affect the total number of women in those ages. The mean length of generation is 28 to 29 years. 13

Rele’s Technique Procedure: Although the coefficients were established based on stable population theory, the relationships can be used in actual populations. Once the gross reproduction rates (GRR) have been derived, they are converted to total fertility rates by calculating the ratio of the number of total births to female births, and multiplying this ratio by the GRR. 14

Rele’s Technique Advantages: The technique requires little information – an age distribution and an estimated life expectancy at birth. All modern censuses provide information on the population by age and sex. The technique is not particularly sensitive to the level of mortality, and so errors in estimating the life expectancy at birth have little effect on the estimate of the GRR. 15

Rele’s Technique Limitations: Differential undercount in the age groups used will produce biased results. If children are undercounted to a higher degree than women, which is often the case, the result will be an underestimation of the TFR. Likewise, if children age 0 to 4 are undercounted to a higher degree than those age 5 to 9, the estimates may imply a spurious decline in fertility or a larger decline than is actually true. 16

Rele’s Technique Limitations: The departure of the population from stable conditions will produce biased estimates. Changing fertility and a high level of migration will have a greater impact than changing mortality. Strong age misreporting will affect the results. 17

References Arriaga, Eduardo E. and Associates. 1994. Population Analysis with Microcomputers, Volume 1. Washington, D. C. : U. S. Bureau of the Census, U. S. Agency for International Development, and United Nations Population Fund, chapter IV. Rele, J. R. 1967. “Fertility Analysis Through Extension of Stable Population Concepts. ” Institute of International Studies, University of California, Berkeley, Population Monograph Series, No. 2. _____. 1989. “ 70 Years of Fertility Change in Korea: New Estimates from 1916 to 1985, ” Asia and Pacific Population Forum, vol. 3, no. 2: 29 -53. 18