Child survivorship estimation: methods and data analysis

"The past 20 years have seen extensive elaboration, refinement, and application of the original Brass method for estimating infant and child mortality from child survivorship data. This experience has confirmed the overall usefulness of the methods beyond question, but it has also shown that...estimates must be analyzed in relation to other relevant information before useful conclusions about the level and trend of mortality can be drawn.... This article aims to illustrate the importance of data analysis through a series of examples, including data for the Eastern Malaysian state of Sarawak, Mexico, Thailand, and Indonesia. Specific maneuvers include plotting completed parity distributions and 'time-plotting' mean numbers of children ever born from successive censuses. A substantive conclusion of general interest is that data for older women are not so widely defective as generally supposed."     

Child Mortality in Beirut: Six Indirect Estimates Based on Data Collected at the Time of a Birth

Six indirect techniques for estimating child mortality were applied to information on survival status of all children ever born, provided by a sample of 1,252 women delivered at the Maternity Unit of the American University of Beirut Hospital. The results were compared, using as a reference the estimate derived from partial birth histories (ages of surviving children, ages at death of children who have died). Their dispersion is minimal for the probability of dying between birth and fifth birthday, estimated at 48 per thousand. For each mother, the ratio of the observed number of children who have died, to that expected given the lengths of exposure of the children to the risk of mortality, was used as a dependent variable in a multiple regression analysis. Educational level of mothers had a significant effect, but not occupational status of father, religion or consanguinuity.     

Subnational Non-Hispanic White Natural Decrease in the United States

This research is the first to examine the prevalence and dynamics of non-Hispanic white natural decrease in fine scale subregional units of the United States. In 2015, more non-Hispanic Whites died than were born in 65 percent of the US counties. This is the highest incidence of non-Hispanic white natural decrease ever reported. It results from a complex interaction among fertility, mortality, and migration over a protracted period. Spatial regression is used to identify three critical variables (over-65 population, child–women ratio, and women of childbearing age) that are the immediate demographic causes of this natural decrease. The timely, factual information in this paper provides a demographic context for analysis of the social, political, and policy implications of this emergent demographic phenomenon.     

Child mortality and reproductive behaviour in German village populations in the past: A micro-level analysis of the replacement effect

Abstract Reproductive histories of couples married during the eighteenth and nineteenth centuries in a sample of 14 German villages are analysed in order to answer several questions regarding the relationship between child mortality and reproductive behaviour. An effort is made through selection of cases and use of multiple classification analysis to eliminate or control non-volitional or otherwise confounding influences on the relationship between a couple's experience with child mortality and their fertility. The results do not provide a decisive answer to the question of whether, under a regime of otherwise presumed natural fertility, previous experience of child mortality affected subsequent reproductive behaviour. The evidence was much clearer in indicating that behaviour consistent with replacement efforts emerged or strengthened as family limitation spread. Finally, the results indicated that though it was not necessary for overall child mortality to decline before family limitation practices were adopted, couples with the most favourable child mortality experience were most likely to practise family limitation and to reduce their fertility. Child mortality appeared at least to impede, if not totally prevent, efforts to reduce the number of children ever born or to cease childbearing at an earlier age or at a given parity.     

Changes in fertility patterns can improve child survival in Southeast Asia

This analysis of 1988 Philippine Demographic Survey data provides information on the direct and indirect effects of several major determinants of childhood mortality in the Philippines. Data are compared to rates in Indonesia and Thailand. The odds of infant mortality in the Philippines are reduced by 39% by spacing children more than two years apart. This finding is significant because infant mortality rates have not declined over the past 20 years. Child survival is related to the number of children in the family, the spacing of the children, the mother's age and education, and the risks of malnutrition and infection. Directs effects on child survival are related to infant survival status of the preceding child and the length of the preceding birth interval, while key indirect or background variables are maternal age and education, birth order, and place of residence. The two-stage causation model is tested with data on 13,716 ever married women aged 15-49 years and 20,015 index children born between January 1977 and February 1987. Results in the Philippine confirm that maternal age, birth order, mortality of the previous child, and maternal education are directly related to birth interval, while mortality of the previous child, birth order, and maternal educational status are directly related to infant mortality. Thailand, Indonesia, and the Philippines all show similar explanatory factors that directly influence infant mortality. The survival status of the preceding child is the most important predictor in all three countries and is particularly strong in Thailand. This factor acts through the limited time interval for rejuvenation of mother's body, nutritional deficiencies, and transmission of infectious disease among siblings. The conclusion is that poor environmental conditions increase vulnerability to illness and death. There are 133% greater odds of having a short birth interval among young urban women than among older rural women. There is a 29% increase in odds for second parity births compared to third or higher order parities. Maternal education is a strong predictor of infant survival only in the Philippines and Indonesia. Adolescent pregnancy is a risk only in Indonesia. Socioeconomic factors are not as important as birth interval, birth order, and maternal education in determining survival status.     

Demographic Determinants of Infant and Early Child Mortality: A Comparative Analysis

In this paper we examine the relative importance of a number of demographic determinants of infant and early child mortality using information from 39 World Fertility Survey countries. We include sex of the child, age of the mother at the time of the birth, birth order, mother's educational level and a number of indicators of spacing of adjacent births among the correlates of chances of survival for children below the age of five years. Mortality of firstborn children and of those born to teenage mothers is shown to be higher than average; that of later children and those of older mothers was not much higher than average, once other factors are controlled. Effects of poor birth-spacing persist even after other factors have been controlled, and are similar where a sib was born during the two years preceding the birth of the child, regardless of the survival status of that sib; however, mortality was higher when that sib had died, due to increased familial risks of mortality. Rapid subsequent births also raise mortality for their earlier sibs. The findings are generally remarkably consistent in a wide range of countries and associated mortality conditions, although attention is drawn to a few interesting geographically clustered exceptions which deserve further investigation. The study leaves little room for doubt that poor child-spacing is clearly linked to decreased survival chances.     

Birth Spacing and Child Health Trajectories

Using longitudinal data on a cohort of over 4,000 children from four low- and middle-income countries, we document the association between birth spacing and child growth trajectories. We find declines in child height at age 1 among children who are born within three years of an older sibling. However, we also observe catch-up growth for closely spaced children as they age. We find no evidence that catch-up growth is driven by remedial health investments after birth, suggesting substitutability in underlying biological processes. We also find that very widely spaced children (preceding birth interval of more than seven years) are similar in height at age 1 as children who are spaced three to seven years apart, but outgrow their more closely spaced counterparts as they age. However, further sibling comparisons suggest that the growth premium that is observed for very widely spaced children may be driven by unobserved confounding factors.     

Models for the estimation of the probability of dying between birth and exact ages of early childhood

Abstract This paper develops two models, each of which is designed to estimate the probability of surviving from birth to selected exact ages of early childhood: namely ages two, three and five. The models are designed for use in areas with deficient registration systems. They require, as input, statistics which can be derived from retrospective data supplied by census or survey respondents. The first model, the age model, converts statistics on the proportion dead of children ever born to women in age groups 20-24, 25-29 and 30-34 into estimates of q2, q3 and q5. The second model, the marriage model, converts statistics on the proportion dead of children ever born to women of five-year marriage duration intervals into these estimates. The models can be used independently or simultaneously. These models were developed from data generated by a large number of empirical fertility and mortality schedules. Regression analysis was used to determine the parameter values of the relationships specified, and several sets of equations for estimating values of qa, for a = 2, 3 and 5 comprise the final product of the paper. It should be noted that the conceptual basis for the models was first derived by William Brass. The data generated for the regression analysis provided an opportunity to test the original Brass estimated model. We are able to report that the model performed well over the wide range of fertility and mortality conditions included in the test.     

Measuring the impact of child mortality on fertility: A methodological note

Recent studies of the impact of child mortality on children ever born have used the “replacement factor” to measure mortality. When microlevel data are used, however, use of the replacement factor (or other variables which are nonlinear in the family’s experienced child mortality rate) yields biased coefficients. An alternative model suggests a sequential, rather than a static, decision-making process to relate the decision to have an additional child to the reproductive experience to date. In this case, unbiased coefficients are obtained if the functional form is correctly specified. In the absence of a priori knowledge of the functional form, it is difficult to untangle true effects from spurious ones.     

Reassessing the decline in parent-child old-age coresidence during the twentieth century

The share of the elderly living with an adult child decreased monotonically throughout the twentieth century, while the probability of reaching old age and the number of years lived in old age increased. As a result, the expected number of life-years lived with adult children while in old age may have increased, decreased, or stayed the same. I estimate that the number of life-years lived in old-age coresidence with adult children stayed roughly constant between 1900 and 1940, while the rate of coresidence declined. Life years lived in old-age coresidence then declined substantially between 1940 and 1990. Moreover, the number of life-years lived in old-age coresidence in 1990 would have been roughly half as great as it actually was had there been no improvements in mortality between 1900 and 1990. And if fertility had remained at its 1900 levels, life-years lived in old-age coresidence would have been about 45% higher in 1990 than it actually was. The results imply that analyses of the change in familial assistance to the elderly should also consider changes in mortality.     

Fertility trends in rural Hebei Province,People's Republic of China

Background: The People's Republic of China (PRC) has conducted several different population policies since its establishment. Although fertility has declined dramatically in the past three decades, the degree to which this was the result of the different population policies is still under debate. Purpose: We attempt to evaluate the effect of the different formal population policies conducted in the PRC by looking at the fertility behavior of rural women. Unlike urban women, rural women experienced less social control (in the absence of a work unit) and received fewer benefits from adhering to the one-child policy. Data: The data analyzed were collected from a stratified sample of households from 288 villages in 9 counties of Hebei Province, PRC, between 1996 and 1999. The number of children ever born was reported by 4,168 ever-married women aged 25 and over who had had at least one birth. Findings: Our analysis indicates that the formal population policies of the PRC had little effect on the number of children ever born to rural women in Hebei. These retrospective data, by cohort, indicate consistently declining fertility since the revolution (1949). Limited child bearing was associated with age and the level of education. Controlling for the effect of age and education, women born after 1960, at whom the one-child policy was directed, actually had more children than older women. Conclusions: The Chinese fertility decline, at least as reflected in the experience of rural women in Hebei Province, derived mainly from secular changes in women's access to education and other social resources rather than from the direct effects of population policies.     

Socio-economic Differentials in Infant and Child Mortality during Mortality Decline: England and Wales, 1890–1911

In this paper data from the 1911 Census of the Fertility of Marriage of England and Wales are used to study patterns of mortality decline by socio-economic characteristics, principally the occupation of husband. That census reported data on number of wives, children ever born, and children dead by marriage-duration cohorts for 190 non-overlapping occupations of husband. These results, along with those on number of rooms in the dwelling of the family are used to make indirect estimates of childhood mortality using the techniques described in United Nations, Manual X. These procedures produce values of q(a), the probability of dying before reaching some exact age ‘a’. Estimates for q(2), q(3), q(5), q(10), q(15), and q(20) are derived from data on women married 0–4, 5–9, 10–14, 15–19, 20–24, and 25–29 years, respectively. These estimates can also be dated to a point in the past. These values can also be converted to a corresponding level of a Model West life table, which describes the ‘average’ mortality regime which the children of those women experienced. This furnishes a basis to look at mortality decline for various social classes and occupational groups. Ordinary least squares regressions of the levels of Model West life tables implied by the 1(a) values on time give one measure of mortality decline. Another is the absolute amount of the increase in the level of the Model West life tables from marriage-duration cohort 20–24 years to 0–4 years. The aggregate results indicate that social class in England and Wales during the 1890s and 1900s tended to be related to the speed of mortality decline: childhood mortality declined more rapidly in the higher and more privileged social class groups. But the results were neither nearly as strong nor as regular as those which predicted the level of mortality within any marriage-duration cohort. These outcomes are not particularly sensitive to the three different social-class stratification schemes used: the 1911 English Registrar General's classification; the 1951 English Registrar General's classification; and the 1950 U.S. Census classification. There was also a fairly regular and predictable gradient for the number of rooms in the home: child mortality was higher in families who lived in larger dwellings. Analysis of 190 detailed male occupational groups revealed that considerably more of the variation in mortality levels than of trends could be explained by social-class categories. Between 20 and 40 per cent of variation in mortality trend could be accounted for by social class alone, as opposed to 50 to 80 per cent of mortality levels for different marriage-duration cohorts. Results for a more restricted sample of 116 occupations for which income estimates could be made revealed a similar pattern. In addition, income was virtually unrelated to the pattern of mortality decline, and improvement was more rapid in groups who were more urban. This reflects the role of rapidly improving urban sanitation in the late nineteenth and early twentieth centuries in England. In contrast, income was significantly related to childhood morality levels for various marriage-duration cohorts (with higher income associated with lower mortality), while urbanization was inversely correlated with mortality levels (more urban groups experienced higher mortality). Overall, social class (or occupation group), income, and urbanization were more successful in explaining mortality levels than time trends across occupations, although social class and the extent of urbanization did reasonably well in accounting for trends. Over a longer period, the transition in child mortality was under way by the 1890s, but its pace and timing varied in different occupations and social class groupings. Although absolute differences in infant mortality were reduced after about 1911, relative inequality persisted even as infant and child survival improved for all groups.     

Childhood Mortality Estimates Using the Preceding Birth Technique: Some Applications and Extensions

The Preceding Birth Technique has proved to be a simple and effective way of estimating mortality during the first two years of life when the question on the survival of the preceding child is asked at the time of a subsequent delivery. This article shows first that earlier worries about significant biases in the estimates are unfounded. Amongst the possible sources of bias considered are: the link between birth-interval length and child survival; the effects of birth intervals differing from 30 months; and the contribution of other systematic biases, such as the omission of all reports on the survival of a woman's last child. Results from a trial in the maternity clinics in Bamako, Mali are presented to illustrate both practical aspects of the method and selection effects by age and parity. A strong association between the survival chances of the succession of children born to one mother emerges. The second part of the article describes how the method can be applied more widely by asking the key questions at times other than a subsequent delivery. Adjustment factors are developed to correct for the biases introduced when the sample includes only mothers whose last-born children are alive at interview.Finally, some examples are given of recent use of the question on the survival of the preceding child in household surveys.     

Risk Exposure in Early Life and Mortality at Older Ages: Evidence from Union Army Veterans

This study examines the relation between risk exposures in early life and hazard of mortality among 11,978 Union Army veterans aged 50 and over in 1900. Veterans' risk exposures prior to enlistment—as approximated by birth season, country of birth, residential region, city size, and height at enlistment—significantly influenced their chance of survival after 1900. These effects are robust irrespective of whether socioeconomic well‐being in 1900 has been taken into account; however, they are sensitive to the particular age periods selected for survival analysis. Whereas some of the effects such as being born in Ireland and coming from large cities became apparent in the first decade after 1900 and then dissipated over time, the effects of birth season, being born in Germany, residential region in the United States, and height at enlistment were more salient in the post‐1910 periods. Height at enlistment shows a positive association with risk of mortality in the post‐1910 periods. Compared to corresponding findings from more recent cohorts, the exceptional robustness of the effects of risk exposures prior to enlistment on old‐age mortality among the veterans highlights the harshness of living conditions early in their lives.     

Communicating population and family planning

This thorough look at the change in the American family 1900-1700 finds that 40% of marriages among women now in their late 20s may end in divorce, that the divorce rate is stabilizing, that between 1-4% of unrelated men and women are living together in informal unions (the figure made difficult to obtain by the difficulty in framing the question), that 15 million adults live alone, and that only 67% of children live with their own once-married parents. About 33% of births are premaritally conceived. The median age for mothers at birth of last child has moved downward from 33 years in the early 1900s to about 30 years. Childbearing has declined from 3.9 children per mother in the early 1900s to 2.5. The period of childbearing has been compressed to about 7 years, between ages 23-30. 10% of remarried women's children are born between marriages. 50% of pregnancies end in abortion. It was found that persons who had completed an educational level, whether it be high school or college, generally had more stable marriages; those who had not completed a level were more likely to get divorced. Despite changes in lifestyle, however, some typical family situations are experienced by most Americans. 2 of 3 marriages will last until death of 1 of the partners and most young women questioned in census surveys expect 2 children.     

Bluestockings, spinsters and pedagogues: Women college graduates, 1865-1910

Summary This article is a study of the demographic behaviour of women college graduates in late nineteenth and early twentieth-century America. The nuptiality and fertility patterns of this group of highly educated women are described, and several explanations of their 'unusual' behaviour are evaluated. Marriage rates of women college graduates declined during the second half of the nineteenth century, even as more women attended college. Only about half the women graduating during the 1890s ever married. Still, the number of children ever born per alumna only varied between 1.0 and 1.5 for the graduation classes of 1865 to 1910. An explanation based on changing labour market opportunities for educated men and women best explains this population's demographic patterns over time as well as their deviations from those of other women in their birth cohort.     

An alternative sampling strategy for obtaining child health data in a reproductive health survey

Retrospective demographic surveys typically collect substantial information about child health. This information is often collected for all children born during a specified period. For women with several young children, the interview can become quite long. To shorten the interview, some surveys have asked child health questions only for the last child born. However, data on the last birth may be biased because last children have a younger age distribution and have longer subsequent birth intervals than does the average child. In this paper, we propose an alternative approach to collecting child health data - that child health questions be asked only for a child chosen randomly from among the respondent's children younger than age five. This alternative has the advantage of keeping the interview shorter but does not lead to biased information.Abbreviations DHS Demographic and Health Surveys     

A flexible two-dimensional mortality model for use in indirect estimation

Mortality estimates for many populations are derived using model life tables, which describe typical age patterns of human mortality. We propose a new system of model life tables as a means of improving the quality and transparency of such estimates. A flexible two-dimensional model was fitted to a collection of life tables from the Human Mortality Database. The model can be used to estimate full life tables given one or two pieces of information: child mortality only, or child and adult mortality. Using life tables from a variety of sources, we have compared the performance of new and old methods. The new model outperforms the Coale-Demeny and UN model life tables. Estimation errors are similar to those produced by the modified Brass logit procedure. The proposed model is better suited to the practical needs of mortality estimation, since both input parameters are continuous yet the second one is optional.     

Birth and death in cities in the developing world

City dwellers in Sub-Saharan Africa have increased roughly 600% in the last 35 years. Throughout the developing world, cities have expanded at a rate that has far outpaced rural population growth. Extensive data document lower fertility and mortality rates in cities than in rural regions. But slums, shantytowns, and squatters' settlements proliferate in many large cities. Martin Brockerhoff studies the reproductive and health consequences of urban growth, with an emphasis on maternal and child health. Brockerhoff reports that child mortality rates in large cities are highest among children born to mothers who recently migrated from rural areas or who live in low-quality housing. Children born in large cities have about a 30% higher risk of dying before they reach the age of 5 than those born in smaller cities. Despite this, children born to migrant mothers who have lived in a city for about a year have much better survival chances than children born in rural areas to nonmigrant mothers and children born to migrant mothers before or shortly after migration. Migration in developing countries as a whole has saved millions of children's lives. The apparent benefits experienced in the 1980s may not occur in the future, as cities continue to grow and municipal governments confront an overwhelming need for housing, jobs, and services. Another benefit is that fertility rates in African cities fell by about 1 birth per woman as a result of female migration from villages to towns in the 1980s and early 1990s. There will be an increasing need for donors and governments to concentrate family planning, reproductive health, child survival, and social services in cities, particularly in Sub-Saharan Africa, because there child mortality decline has been unexpectedly slow, overall fertility decline is not yet apparent in most countries, and levels of migration to cities are anticipated to remain high.     

Magic Moment? Maternal Marriage for Children Born Out of Wedlock

To test the existence of the “magic moment” for parental marriage immediately post-birth and to inform policies that preferentially encourage biological over stepparent marriage, this study estimates the incidence and stability of maternal marriage for children born out of wedlock. Data came from the National Survey of Family Growth on 5,255 children born nonmaritally. By age 15, 29 % of children born nonmaritally experienced a biological-father marriage, and 36 % experienced a stepfather marriage. Stepfather marriages occurred much later in a child’s life—one-half occurred after the child turned age 7—and had one-third higher odds of dissolution. Children born to black mothers had qualitatively different maternal marriage experiences than children born to white or Hispanic mothers, with less biological-parent marriage and higher incidences of divorce. Findings support the existence of the magic moment and demonstrate that biological marriages were more enduring than stepfather marriages. Yet relatively few children born out of wedlock experienced stable, biological-parent marriages as envisioned by marriage promotion programs.