Revisiting the fertility transition: long term perspectives

The purpose of this study is to estimate the effect of spatial diffusion on fertility behaviour changes in Belgium during the First Demographic Transition at a fine-scale geographic level. The analyses lie in a new data series from 1886 to 1913 providing yearly fertility indicators at the municipality level. Firstly, this data series will be used in a preliminary analysis of the evolution of fertility behaviour in order to produce a first overview of spatial patterns, diffusion effects and barriers to diffusion. This preliminary analysis is expected to reveal local particularities of fertility changes that are undetectable at higher geographic levels. Secondly, the yearly data series will be used in event-history models in order to estimate neighbourhood effects among municipalities. These models will test whether the presence of contiguous municipalities j that experienced a fertility change at time t−1 has an effect on the probability, for a municipality i, of experiencing the same change at time t. These models will also attempt to estimate the barrier effect of the Belgian linguistic border.

This paper examined age-pattern of childbearing in terms of [de]compression in age-specific
fertility rates in India and major states. This study followed a two part methodology: first,
compression in ages-specific fertility rates are estimated in terms of skewness and standard
deviation. Second, age bands in which compression in childbearing taking place are identified.
The results reveal that age-pattern of childbearing is highly associated with total fertility rate
levels: with decline in total fertility rates, the age-specific fertility rates are compressing, such
compression is mostly at peak ages (20-30 years). However, there is huge state level and ruralurban
disparity in age-pattern of childbearing by levels of total fertility rates. The state with
greater decline in fertility or currently with low fertility rate show evidence of compression in
age-specific fertility rates while states with lower decline in fertility or currently with high
fertility show greater dispersion in age-specific fertility rates. Urban areas evident for
compression in age-specific fertility rates while rural areas still shows dispersion in age-specific
fertility rates. Compressions in age-pattern of childbearing yet occur in high fertility rates and
rural areas of India.

The interaction of migration with fertility and mortality is key for demographic dynamics at subnational and local levels. Despite this, standard measures of mortality and reproduction treat migration only as a nuisance to be removed, without ever assessing the extent to which processes are affected by migration. Indicators of replacement that incorporate the effect of migration like the Birth Replacement Ratio, have the limitation of not being defined at the age-specific level. We estimate new cohort and period measures of mortality and reproduction that include the interaction with migration for the set of Spanish provinces in the period 1858-2010. The new measures only require vital registration data classified by age, and provide age-specific estimates that incorporate the effects of migration. When migration is low, the measures can be regarded as indirect estimates of mortality and reproduction that do not require population denominators. When migration is high, they should be regarded as the results of the interaction of demographic processes in that particular population. Given the large provincial contrasts in fertility, mortality and migration, it is possible to characterize the role of migration patterns in the vital events registered in the provinces over a period that covers most of demographic transition.

The aim of the study is to highlight some key points of Italian fertility decline in the first decades of the twentieth century, by using a micro longitudinal approach. The research is based on new individual-longitudinal data and life-course histories from Granarolo, an Italian municipality bordering the town of Bologna. By using Cox models, hazards of having a child are estimated for married women between 1900 and 1940. Individual fertility propensity by marriage cohort and period is assessed, controlling for differences in socio-demographic characteristics and describing the changing reproductive strategies. Moving from one marriage cohort to the next one, the hazard of having a first child progressively increased. On the contrary, probabilities of giving birth progressively fell when women reached higher parities. Fertility continued to decrease even if the Fascist Regime promoted a pro-births campaign against fertility decline. Clear differences in fertility were also visible between socioeconomic groups: the lower fertility of upper classes is confirmed even in a rural community such as Granarolo, whereas sharecroppers’ families remained tied to their own traditional higher fertility.

Our aim is to explore socioeconomic fertility differentials in an industrializing community; to gain insight about the details and discuss possible mechanisms. The study starts well before industrialization and finishes at the end of the transition. We use longitudinal individual-level data from the Scanian Economic-Demographic Database, which contains demographic as well as socioeconomic information, including occupation, landholding and income. In the analysis we use hazard regressions with shared frailty at the family level. The transition involved not only parity-specific stopping but also spacing. While the upper social strata had higher fertility prior to the transition, they started to control their fertility earlier, by the 1880s, and also more consistently. Farmers, the middle class and skilled workers followed in the decades after, and unskilled workers with some additional delay. These findings are inconsistent with several of the major explanations in the literature, such as mortality decline, increased female labor force participation and a quantity-quality trade-off, but consistent with an innovation process where new ideas and attitudes about family limitation spread from the elite to other social groups.

In the developing world it is expected that half of the future population growth will be caused by its young population age-structure, an aspect of population dynamics that is referred to as population momentum. Less discussed is the amount of population decline that will be observed in the future in developed countries with its “negative momentum” caused by its old population age-structure. Population momentum summarizes the relation between a population age-structure and the age-structure resulting from its equivalent population under stability (growth equal to zero). Since mortality and fertility change permanently also the population momentum corresponding for each time is different. However, little work has been done to summarize time trends of this measure. We study the trends in time of population momentum for countries with available period and cohort, fertility and mortality information. Our preliminary results show clear distinction between three periods in population momentum, distinguish by their levels of fertility and mortality: high-high, high-low, and low-low, for fertility and mortality respectively. In recent years most developed countries have entered into the low-low stage and more than a decade of negative population momentum.

A negative association between the number of children in a family and the heights of the children has been shown many times for 20th century populations. The causes behind this association are not known but could include resource scarcity in large families or that families with different numbers of children are different also in other, observable or unobservable, ways. This study investigates the causes behind the negative association between family size and child heights by examining the association among men born 1797-1950. This is done using longitudinal, individual level data on socioeconomic background, household structure, family relations and heights of conscripts. The results show that the causes behind the negative association are manifold and complex, and their relative importance having changed over time. Children with many siblings were shorter than others, at least from mid-19th century onwards. Resource dilution within families is probably one of the explanations for this negative association. But the association seems to have been different in landed families as compared to others, and also different in families with unusually few or many children. The explanations needed to explain the negative association between sibship size and child outcomes differ depending on the societal and historical context.

This paper provides research-based evidence to examine the association between family formation, fertility transition, and cultural changes from a demographic perspective. The evidence of this paper is obtained from a country at which one of the most spectacular falls in birth rate ever experienced in human history has occurred: Iran. The country's fundamental socio-cultural changes over the past decades have made it as a unique ‘human and social laboratory’ to survey their consequences on family formation characteristics such as marriage and fertility patterns. As this paper analyses family formation characteristics in the varying socio-cultural circumstances, it also provides evidence for the long-standing debate as to whether culture and religion itself or other factors explain family formation characteristics (such as early marriage, high incidence of marriage and high fertility for women) in these sorts of religious and cultural contexts.

Demographers demonstrate an increasing interest for studying parental gender preferences in developing countries and more developed societies. They find an association between the sex-composition of living offspring and the probability of having an additional child. In history, parents’ gender preferences have proven difficult to verify. This study makes use of John Knodel’s German village genealogies to obtain knowledge on this issue during a period of fertility transition, 1825-1900. Couples at first marriage who gave birth to minimum four children are targeted. Event history analyses (Cox regression models) of couples’ duration and propensity to progress to fifth parity, helps us to test if the probability to have additional children was influenced by the sex-composition of surviving children at lower parities. It appears that preferences for son(s) did influence parents’ reproductive behavior, as those having only girls experienced the highest transition rates to fifth parity. However, couples married from approximately 1870 onward started to exhibit a fertility behavior consistent with the desire to have at least one surviving boy and girl. That the gender preferences became more symmetrical already during the fertility decline we view as an surprisingly early move toward a modern European pattern.

The transitional decline of fertility in Italy has never been studied using micro-data, with the exception of a few small areas. For the first time, we use individual retrospective fertility data collected for 20% of married women aged 40 or older interviewed in the 1971 census in the Veneto Region (North-East Italy), a “late-comer” area in the context of Western European fertility decline (TFR=5.0 in 1871 and 1921, 2.5 in 1951). In order to consider broad explanations of fertility decline, we combine individual retrospective data with other information available at two territorial levels (65 districts or large cities and 582 municipalities), using a three-level clustered regression model (district, municipality, woman). We find that: (1) even if the (few) university-educated women born in the last decades of the 19th century already had a TFR around two, this value is not seen among women with low levels of education until those born fifty years later; (2) the link between fertility and secularization strengthens cohort after cohort, whereas the connections between fertility and industrialization and fertility and urbanization weaken.