Joseph Molitoris | University of Copenhagen (original) (raw)

Uploads

Papers by Joseph Molitoris

European Journal of Population

Perspectives on Sexual and Reproductive Health

This study investigates the association between individual-level characteristics and the risk of ... more This study investigates the association between individual-level characteristics and the risk of having an unmet need for contraception in the United States between 2002 and 2017 for women who were sexually active, not pregnant or postpartum, fecund, and wanted no more children. Using data from the National Survey of Family Growth, logistic regression models are used to estimate the association between demographic, socioeconomic, and reproductive characteristics and the probability of having an unmet need for any contraception and for modern methods. The results show that 6.8% of the women at risk did not use any form of contraception and 12.1% did not use any modern method. Women who were black, Catholic, insured by Medicaid or uninsured, nulliparous, and who had not used contraception at their first sex had the greatest odds of having an unmet need, suggesting that specific groups are disproportionately vulnerable to unwanted pregnancy in the United States.

Studies in Family Planning

SSRN Electronic Journal, 2000

The timings of historical fertility transitions in different regions are well understood by demog... more The timings of historical fertility transitions in different regions are well understood by demographers, but much less is known regarding their specific features and causes. In the study reported in this paper, we used longitudinal micro-level data for five local populations in Europe and North America to analyse the
relationship between socio-economic status and fertility during the fertility transition. Using comparable analytical models and class schemes for each population, we examined the changing socio-economic differences in marital fertility and related these to common theories on fertility behaviour. Our results do not provide support for the hypothesis of universally high fertility among the upper classes in pretransitional society, but do support the idea that the upper classes acted as forerunners by reducing their fertility before other groups. Farmers and unskilled workers were the latest to start limiting their fertility. Apart from these similarities, patterns of class differences in fertility varied significantly between populations.

This work combines economic and demographic data to examine inequality of living standards in Sto... more This work combines economic and demographic data to examine inequality of living standards in Stockholm at the turn of the twentieth century. Using a longitudinal population register with occupational information, we utilize event-history models to show that despite absolute decreases in mortality, relative differences between socioeconomic groups remained virtually constant. The results also show that child mortality continued to be sensitive to short-term fluctuations in wages and that there were no socioeconomic differences in this response. We argue that the persistent inequality in living standards was possibly due to differences in residential patterns and nutrition.

The western fertility decline is arguably the most significant demographic change to have occurre... more The western fertility decline is arguably the most significant demographic change to have occurred in the past 200 years, yet its causes and processes are still shrouded in ambiguity due to a lack of individual-level longitudinal data. A growing body of research has helped improve our understanding of the decline’s causes by examining the development of socioeconomic differences in fertility using historical micro-data, but these have largely only considered rural areas where fertility was generally slower to decline. This article contributes to the literature by utilizing individual-level data
from the Roteman Database for Stockholm, Sweden between 1878 and 1926 to examine the association of socioeconomic status and fertility and the adoption of stopping behaviour during the city’s transition. Using piecewise constant hazard models and logistic regression, we find that a clear class pattern arises in which the
elite were early practitioners of fertility control, followed by the working classes. As the transition unfolded, socioeconomic differences in stopping behaviour disappeared and overall fertility differentials were also minimized, both of them being consistent
with patterns observed in rural populations. The implications of these findings for major explanations of the decline are discussed in the concluding section.

A negative association between birth interval length and infant and child mortality has been cons... more A negative association between birth interval length and infant and child mortality has been consistently identified in modern developing countries. The reasons for this association are unclear, however. Leading hypotheses linking interval length to mortality fall into four broad categories—sibling competition, maternal depletion, infection transmission, or unobserved maternal factors—but none has received overwhelming support. Using data from Stockholm between 1878 and 1926, this study identifies trends in the relationship over time, controlling for unobserved maternal heterogeneity, and exploiting sibling deaths to better understand the mechanisms at work. Results show that the association disappeared over time as infectious disease mortality fell and that deaths of previous siblings during the postnatal period disproportionately tended to increase the risk of dying among index children born after short intervals. These findings strongly suggest the relationship is related to the transmission of disease between closely spaced siblings.

BACKGROUND The decline of child mortality during the late 19th century is one of the most signifi... more BACKGROUND
The decline of child mortality during the late 19th century is one of the most significant demographic changes in human history. However, there is evidence suggesting that the substantial reductions in mortality during the era did little to reduce mortality inequality between socioeconomic groups.

OBJECTIVE
The aim of this study is to examine the development of socioeconomic inequalities in cause-specific infant and child mortality during Stockholm’s demographic transition.

METHODS
Using an individual-level longitudinal population register for Stockholm, Sweden, between 1878 and 1926, I estimate Cox proportional hazards models to study how inequality in cause-specific hazards of dying from six categories of causes varied over
time. The categories included are 1) airborne and 2) food and waterborne infectious diseases, 3) other infectious diseases, 4) noninfectious diseases and accidents, 5) perinatal causes, and 6) unspecified causes.

RESULTS
The results show that class differentials in nearly all causes of death converged during the demographic transition. The only exception was the airborne infectious disease category, for which the gap between white-collar and unskilled blue-collar workers
widened over time.

CONCLUSIONS
The results demonstrate that, even in a context of falling mortality and a changing epidemiological environment, higher socioeconomic groups were able to maintain a health advantage for their children by reducing their risks of dying from airborne disease to a greater extent than other groups. Potential explanations for these patterns
are suggested, as well as suggestions for future research.

European Journal of Population

Perspectives on Sexual and Reproductive Health

This study investigates the association between individual-level characteristics and the risk of ... more This study investigates the association between individual-level characteristics and the risk of having an unmet need for contraception in the United States between 2002 and 2017 for women who were sexually active, not pregnant or postpartum, fecund, and wanted no more children. Using data from the National Survey of Family Growth, logistic regression models are used to estimate the association between demographic, socioeconomic, and reproductive characteristics and the probability of having an unmet need for any contraception and for modern methods. The results show that 6.8% of the women at risk did not use any form of contraception and 12.1% did not use any modern method. Women who were black, Catholic, insured by Medicaid or uninsured, nulliparous, and who had not used contraception at their first sex had the greatest odds of having an unmet need, suggesting that specific groups are disproportionately vulnerable to unwanted pregnancy in the United States.

Studies in Family Planning

SSRN Electronic Journal, 2000

The timings of historical fertility transitions in different regions are well understood by demog... more The timings of historical fertility transitions in different regions are well understood by demographers, but much less is known regarding their specific features and causes. In the study reported in this paper, we used longitudinal micro-level data for five local populations in Europe and North America to analyse the
relationship between socio-economic status and fertility during the fertility transition. Using comparable analytical models and class schemes for each population, we examined the changing socio-economic differences in marital fertility and related these to common theories on fertility behaviour. Our results do not provide support for the hypothesis of universally high fertility among the upper classes in pretransitional society, but do support the idea that the upper classes acted as forerunners by reducing their fertility before other groups. Farmers and unskilled workers were the latest to start limiting their fertility. Apart from these similarities, patterns of class differences in fertility varied significantly between populations.

This work combines economic and demographic data to examine inequality of living standards in Sto... more This work combines economic and demographic data to examine inequality of living standards in Stockholm at the turn of the twentieth century. Using a longitudinal population register with occupational information, we utilize event-history models to show that despite absolute decreases in mortality, relative differences between socioeconomic groups remained virtually constant. The results also show that child mortality continued to be sensitive to short-term fluctuations in wages and that there were no socioeconomic differences in this response. We argue that the persistent inequality in living standards was possibly due to differences in residential patterns and nutrition.

The western fertility decline is arguably the most significant demographic change to have occurre... more The western fertility decline is arguably the most significant demographic change to have occurred in the past 200 years, yet its causes and processes are still shrouded in ambiguity due to a lack of individual-level longitudinal data. A growing body of research has helped improve our understanding of the decline’s causes by examining the development of socioeconomic differences in fertility using historical micro-data, but these have largely only considered rural areas where fertility was generally slower to decline. This article contributes to the literature by utilizing individual-level data
from the Roteman Database for Stockholm, Sweden between 1878 and 1926 to examine the association of socioeconomic status and fertility and the adoption of stopping behaviour during the city’s transition. Using piecewise constant hazard models and logistic regression, we find that a clear class pattern arises in which the
elite were early practitioners of fertility control, followed by the working classes. As the transition unfolded, socioeconomic differences in stopping behaviour disappeared and overall fertility differentials were also minimized, both of them being consistent
with patterns observed in rural populations. The implications of these findings for major explanations of the decline are discussed in the concluding section.

A negative association between birth interval length and infant and child mortality has been cons... more A negative association between birth interval length and infant and child mortality has been consistently identified in modern developing countries. The reasons for this association are unclear, however. Leading hypotheses linking interval length to mortality fall into four broad categories—sibling competition, maternal depletion, infection transmission, or unobserved maternal factors—but none has received overwhelming support. Using data from Stockholm between 1878 and 1926, this study identifies trends in the relationship over time, controlling for unobserved maternal heterogeneity, and exploiting sibling deaths to better understand the mechanisms at work. Results show that the association disappeared over time as infectious disease mortality fell and that deaths of previous siblings during the postnatal period disproportionately tended to increase the risk of dying among index children born after short intervals. These findings strongly suggest the relationship is related to the transmission of disease between closely spaced siblings.

BACKGROUND The decline of child mortality during the late 19th century is one of the most signifi... more BACKGROUND
The decline of child mortality during the late 19th century is one of the most significant demographic changes in human history. However, there is evidence suggesting that the substantial reductions in mortality during the era did little to reduce mortality inequality between socioeconomic groups.

OBJECTIVE
The aim of this study is to examine the development of socioeconomic inequalities in cause-specific infant and child mortality during Stockholm’s demographic transition.

METHODS
Using an individual-level longitudinal population register for Stockholm, Sweden, between 1878 and 1926, I estimate Cox proportional hazards models to study how inequality in cause-specific hazards of dying from six categories of causes varied over
time. The categories included are 1) airborne and 2) food and waterborne infectious diseases, 3) other infectious diseases, 4) noninfectious diseases and accidents, 5) perinatal causes, and 6) unspecified causes.

RESULTS
The results show that class differentials in nearly all causes of death converged during the demographic transition. The only exception was the airborne infectious disease category, for which the gap between white-collar and unskilled blue-collar workers
widened over time.

CONCLUSIONS
The results demonstrate that, even in a context of falling mortality and a changing epidemiological environment, higher socioeconomic groups were able to maintain a health advantage for their children by reducing their risks of dying from airborne disease to a greater extent than other groups. Potential explanations for these patterns
are suggested, as well as suggestions for future research.