Birth Spacing and Child Survival: A Cautionary Note Regarding the Evidence from the WFS

In this paper results recently published in this journal concerning the influence of birthspacing on child survival are reviewed, and two reasons are suggested why the statistical associations observed in the data collected in the World Fertility Survey may be, at least in part, spurious. Where reporting is poor, it is argued that errors may often be responsible for a substantial portion of the association between interval length and mortality, and also for the apparent influence of the death of one child on the survival prospects of a younger sibling. Where the reported dates of birth and death are reliable but contraceptive prevalence is high, there is the problem that contraceptive practice is apt to be closely associated with other behaviour likely to have a direct effect on mortality.     

The completeness of enumeration in the 1973 census of the population of Colombia

The effort is made to determine the true size and distribution by age and sex of the population of the Republic of Colombia in October 1973. After initially arriving at estimates of the levels of fertility and mortality during the intercensal period and then correcting the 1964 census population for age misreporting and selective undernumeration of males, a hypothetical populaiton corresponding to October 1973 is constructed. Comparing the constructed population with the population observed in the census yelds an estimate of completeness of enumeration in 1973 that is relative to the enumeration of females in 1964. This estimate is obtained under the assumption that net international migration during the period was of negligible importance. As there is reason to believe that this is not a valid assumption and upon examining the limited amount of evidence available, speculaitons are made concerning the amount of net out-migration to have occurred during the 1964-1973 period and the size of the coresponding modificaiton in the estimate of completeness of enumeration. After adjusting for underenumeration of males in 1964 and neglecting the impact of international migration, a theoretical 1973 census population of 23,201,000 was estimated. Apart from the total number of people enumerated, the information that was analyzed from the advance sample appears to be of good quality, at least in relation to prior censuses. The estimates of fertility and mortality reveal an important decline in Colombian fertility. By coming up with separate estimates of infant and childhood and adult mortality, it has been possible to shed new light on the shape and the level of mortality in Colombia. The new Brass method for estimating adult mortality provides reliable results even when mortality has been declining, and there are recognizable distortions in the distribution of the population by age.     

Socioeconomic Determinants of White and Black Males’ Life Expectancy Differentials, 1980

Epidemiological transition theory suggests that two population existing under disparate socioeconomic conditions would have different life expectancies as the result of cause-of-death differences. The effect of racial socioeconomic differentials on the total racial life expectancy differential are examined as they act through specific cause-of-death differentials. Results suggest that residential isolation of blacks has a strong effect on the total life expectancy differential as it acts through the racial homicide differential. The racial unemployment difference also has a strong effect on the total differential as it acts through the racial heart disease differential. Implications of the findings for reducing life expectancy differentials are discussed.     

Number and mean age of acceptors and users as related to programme duration

Abstract A decline in annual acceptances in a contraceptive programme is open to multiple interpretation. The decline may reflect an adverse change of attitudes and therefore a lowered disposition to accept the method among couples of the population generally. Alternatively, if the population has an irregular age distribution, the decline may mean only that the numbers of couples of prime contraceptive age have temporarily dropped. Thirdly, if the rate of previous acceptance has been high, the main cause of a slackening in new acceptances may simply be that previous acceptances have partially depleted the pool of eligible couples. Fewer eligible couples are left than previously as a consequence of past acceptance, removing couples interested in contraception faster than the family-building process can create new ones.     

A computer model of family building based on expected values

A computerized probability model of family building is described. Called FERMOD, the model is designed to follow the changing distribution of children ever born and birth intervals of a large homogeneous population as it moves through the reproductive period. Use of the model presupposes assumptions about such factors as risks of fetal wastage, lengths of pregnancy, and postdelivery amenorrhea conditional to outcome of pregnancy, length of reproductive period, fecundability, desired family size, preferred birth spacing!, and effectiveness of contraception. Use of FERMOD is illustrated in a limited application to United States fertility. Relations to other models are briefly discussed.     

Capacity constraints and the dynamics of underwriting profits

The presence of an underwriting profit cycle in property/liability insurance has become a stylized fact. Models of this'underwriting cycle'imply that the insurance market is governed by two regimes, as capacity is constrained or not. We apply the smooth transition regression model to insurance industry data for 1934–1993 to test for a regime shift. We find a rapid shift between two distinct regimes with different dynamics. When capacity is not restricted, we find no evidence of a cycle. The cycle is present in periods when capacity is restricted, immediately following World War II and after 1968. The underwriting cycle appears to be a recent phenomenon.     

Repeat abortion

A reanalysis of the repeat abortion experience of New York City residents during July 1, 1970 to June 30, 1972 is undertaken on the basis of a probability model that generates repeat abortion ratios as a function of assumptions about fecundity, contraceptive efficiency, and exposure lengths. Tested are three hypotheses put forward by Daily et al. in a 1973 analysis: (i) the low repeat abortion ratio of .0245 is attributable in part to underreporting of registered induced abortions as repeat ones; (ii) a major part of the rise in repeat abortion ratios, from virtually zero to six percent over four consecutive six-month intervals, is explainable in terms of the rising volume of exposure time to risk of repeat abortion relative to the stream of initial abortions; and (iii) the higher abortion ratios of women in their twenties compared to those of older or younger women is ascribable to “differences in fecundity and intercourse frequency.” Support is found for the first two hypotheses, and a mixed outcome for the third.     

Changes in acceptors' and users' ages: A test of an explanatory mechanism

Summary Acceptors in national family planning programmes are becoming steadily younger, with fewer children. In the present analysis, which makes use of the computerized component projection scheme CONVERSE, it is shown that saturation of the older groups with users has a relatively minor role to play in reducing the mean age of new acceptors. Even where acceptance and continuation rates are high and follow the usual pattern of higher levels for the older women, the build-up of users in the upper age groups is not enough to produce major changes in the age of acceptors. Yet acceptor ages have fallen consistently and sharply in most national programmes, regardless of the general strength, duration, or method mix. The finding that the mechanical effects investigated here play rather a small part in the large empirical declines in acceptor ages suggests the need to investigate other explanations. For this, a critical need is more data than are at present available on trends in age-specific rates of acceptance in national programmes.