Maternal and child health

Einleitung / Introduction

In recent decades, many countries have significantly expanded their universal childcare systems to facilitate maternal labor supply, increase gender equality, and improve family well-being. While a large body of literature shows that these expansions led to increased maternal labor force participation and greater gender equality in many countries, there is little evidence on their impact on maternal well-being and health. The increased availability of childcare could affect mother’s health in different ways: Infectious events in childcare facilities, shifts in maternal time usage towards paid work or self-care, and emotional challenges coming with childcare usage, especially at early stages of life. In this paper, we examine the impact of a substantial expansion of universal early childcare in Germany on the health of mothers. Our primary focus is on mothers' self-assessed health, with additional analysis conducted on various indicators of both mental and physical well-being.

Methode / Method

Using SOEP data from 2006 to 2019 and exploiting the staggered expansion of early childcare availability in German counties following the 2005 childcare reform, we are able to identify the causal effects of childcare availability on maternal health.

Ergebnisse / Results

We estimate a county level fixed effects model and find a highly statistically significant effect of early childcare expansion on mothers' self-assessed health. Specifically, for a ten-percentage point increase in local childcare coverage, mothers' self-assessed health diminishes by 0.173 points on a one to five scale (19% of a standard deviation). These effects are more pronounced for mothers with less education and multiple children. Further analyses indicate that mothers experience adverse effects in both the physical and mental health domains. The increased availability of childcare leads to a decline in physical functioning and an increase in mental or emotional difficulties. Moreover, mothers are more inclined to report that they need to curtail their social activities due to health issues and express lower satisfaction with their overall health. Furthermore, we find no effects on mothers’ partners and women without children.

Zusammenfassung / Conclusion

Combining SOEP with county-level data on early childcare usage, we find negative effects of increased childcare availability on maternal self-assessed health, satisfaction with health and indicators of both mental and physical well-being. The variety of control variables, the robustness of the results and the findings of a placebo check provide support for the hypothesis that we are indeed measuring the effects of early childcare expansion rather than confounding trends.

Authors

Niklas Rott, Universität Augsburg

Marina Krauss, Universität Augsburg

Einleitung / Introduction

The age at which women in Germany become mothers for the first time is increasing. However, fertility determinants that may deteriorate with age, such as health, and their effects on reproductive patterns have received little attention. We examine the effect of women's subjective general health at reproductive age on the conditional probability of entering motherhood, focusing on the interplay between health and age.

Methode / Method

The empirical analysis is based on pooled (SOEPO, PAIRFAM) longitudinal survey data from Germany for the years 1995 to 2021. For women aged 17 to 49, we estimate age-specific hazard rates of the transition to motherhood and analyze whether and how subjective health (SAH) is associated with these age-specific hazards. After starting with a simple and purely descriptive analysis, we estimate a series of more sophisticated models (nonparametric regression, conditioning on covariates, instrumental variables to account for unobserved heterogeneity) to see if the initial descriptive pattern is robust to closing possible channels through which confounders might operate.

Ergebnisse / Results

Estimating the different models yields a remarkably robust pattern of results. For women of advanced age, we see a health gradient such that better health is associated with a higher probability of becoming a mother. The pattern is completely reversed for women of relatively young age, for whom good health is associated with a much lower probability of having a first child. These differences are of substantial size, about minus three percentage points at age 23 and about plus three percentage points at age 36. Since conditioning on observables - such as income and education - does not change the pattern of results, it is most likely not driven by socioeconomic status. This likewise applies to unobserved time-invariant individual heterogeneity, as addressing this issue through IV estimation does not qualitatively change the results either.

Zusammenfassung / Conclusion

In contrast to previous research that has examined the effect of health by assuming that it is homogeneous at each point in reproductive age and has often found no such health effect, our analysis reveals a non-monotonic relationship between health and the hazard of transitioning to motherhood. Because our empirical analysis does not exploit a clean source of exogenous variation in health, we cannot interpret the results in terms of precisely identified age-specific health effects. However, as the pattern of results is robust to controlling for several channels through which possible confounders may operate, the results appear to capture more than just the effects of third factors related to both health and fertility. One possible channel through which general health may directly affect fertility decisions could be that women in poorer health may prefer to have their children early in life, rather than wait until an age when health deficits are likely to become more severe.

Authors

Harald Tauchmann, FAU Erlangen-Nürnberg

Irina Simankova, FAU Erlangen-Nürnberg

Einleitung / Introduction

Ramadan intermittent fasting during pregnancy is known to be associated with various adverse health effects in the offspring. Nevertheless, many pregnant Muslims do fast. Since pregnant women regularly exchange with healthcare professionals, the question arises whether the provision of prenatal consultation on Ramadan during pregnancy affects maternal fasting. The analysis of the effects of prenatal consultations in a real-world (non-experimental) setting is complicated by self-selection into treatment, i.e., it is not random which women receive prenatal consultations.

Methode / Method

This study exploits unique survey data from Mainz, Germany. It includes information on Ramadan behavior during pregnancy as well as interviewees’ interactions with health care professionals on Ramadan fasting during pregnancy. We estimate the effect of prenatal consultation on the zero-truncated number of days fasted using a log-normal hurdle model. This allows us to disentangle effects on the extensive and intensive margin, i.e., whether a woman did fast and if so, for how many days. We control for a large set of socio-demographic characteristics. We test whether our controlled estimates might be biased by self-selection into treatment using the novel Oster (2019) method that assesses the potential impact of unobservable confounders.

Ergebnisse / Results

The share of women who fasted at least one day was 36.5% (mean: 17 fasted days). Among these women, receiving prenatal consultation on Ramadan during pregnancy reduced the number of fasted days by 10.8 days (95% CI: -19.03, -2.56). Results from the Oster method show that this result) is not biased by self-selection into treatment.

Zusammenfassung / Conclusion

This study shows a novel way to exploit real-world evidence in the absence of randomized controlled trials. We also show that prenatal consultation on Ramadan fasting significantly reduces the number of fasted days among pregnant Muslims who intend to fast. With a growing Muslim population in Germany, the issue is likely to become even more relevant in the future. Further studies will contribute to the development of guidelines for prenatal advice on Ramadan during pregnancy, for example in a framework of shared decision-making.

Authors

Paul Witte, Johannes Gutenberg-Universität Mainz

Fabienne Pradella, Johannes Gutenberg-Universität Mainz

Reyn Van Ewijk, Johannes Gutenberg-Universität Mainz

Einleitung / Introduction

Inequalities in access to reproductive health care across Europe is an under-researched area. Reproductive health policies and access vary greatly across European countries. England and Wales rates highly for access to reproductive services, including for facilitating access for vulnerable women and cost of contraceptive medicines. In Great Britain, terminations of pregnancy have been legally available since the 1967 Abortion Act, and Northern Ireland recently passed a law facilitating access to abortion. However, even where abortion care is legally granted as a right, formal and informal barriers are at work that reduce people’s access to safe pregnancy terminations. This includes insufficient service provision, stigmatisation of abortion, and information inaccessibility, among others. Qualitative evidence suggests rurality, a social support network, and English language ability limit access to abortions for women in England and Wales The aim of this research is to explore geographic differences in abortion rates in England and Wales, and how they correlate with the clustering of clinics offering abortion services, controlling for a large set of demographic and socio-economic factors, as well as cultural and geographical characteristics.

Methode / Method

We collected information on abortion clinics active in the years 2017-2021 from the Abortion Statistics for England and Wales, the list of independent abortion providers and those listed in the Care Quality Commission (CQC). To identify local authorities (local government administrative areas) the clinics are located in, we used the CQC directory as well as address information available online. We matched the information on active clinics with data at local authority level on abortion rates, and a large set of factors influencing abortion including socioeconomic and inequality indicators, values, and social norms. Based on this rich dataset, we study the clustering of abortion clinics and the determinants of abortion based on spatial regression models.

Ergebnisse / Results

We find a higher number of abortion clinics in the South of England and Midlands. Rural areas, especially in Wales, have lower access. In addition to access to abortion services, significant factors associated with higher abortion rates include higher mean age, lower education and deprivation levels, gender inequality indicators, and urbanity.

Zusammenfassung / Conclusion

Our preliminary results suggest that there are inequalities in access to abortion services for women in England and Wales. In the next steps, we will further investigate geographical access to abortion services using measures of travel time and access to public transportation.

Authors

Federico Cassara, Vienna University of Economics and Business

Heather Brown, Lancaster University

Sophie Guthmuller, Vienna University of Economics and Business