Discrete-time Event History Models Research Articles

Perceived discrimination and incident dementia among older adults in the U.S.: The buffering role of social relationships.

Recent studies have found that perceived discrimination as a chronic stressor predicts poorer cognitive health. However, little research has investigated how social relationships as potential intervening mechanisms may mitigate or exacerbate this association. Using a nationally representative sample of U.S. older adults, this study examined how the existence and quality of four types of relationships-with a partner, children, other family members, and friends-may modify the impact of perceived discrimination on incident dementia. We analyzed data from the 2006-2016 Health and Retirement Study (N=12,236) using discrete-time event history models with competing risks. We used perceived discrimination, social relationships, and their interactions at the baseline to predict the risk of incident dementia in the follow-ups. Perceived discrimination predicted a higher risk of incident dementia in the follow-ups. Although having a partner or not did not modify this association, partnership support attenuated the negative effects of discrimination on incident dementia. Neither the existence nor quality of relationships with children, other family members, or friends modified the association. Our findings imply that intimate partnership plays a critical role in coping with discrimination and, consequently, influencing the cognitive health of older adults. While perceived discrimination is a significant risk factor for the incidence of dementia, better partnership quality may attenuate this association. Policies that eliminate discrimination and interventions that strengthen intimate partnership may facilitate better cognitive health in late life.

Getting ahead in the social sciences: How parenthood and publishing contribute to gender gaps in academic career advancement.

How do parenthood and publishing contribute to gender gaps in academic career advancement? While extensive research examines the causes of gender disparities in science, technology, engineering, and mathematics (STEM) careers, we know much less about the factors that constrain women's advancement in the social sciences. Combining detailed career- and administrative register data on 976 Danish social scientists in Business and Management, Economics, Political Science, Psychology, and Sociology (5703 person-years) that obtained a PhD degree between 2000 and 2015, we estimate gender differences in attainment of senior research positions and parse out how publication outputs, parenthood and parental leave contribute to these differences. Our approach is advantageous over previous longitudinal studies in that we track the careers and publication outputs of graduates from the outset of their PhD education and match this data with time-sensitive information on each individual's publication activities and family situation. In discrete time-event history models, we observe a ∼24 per cent female disadvantage in advancement likelihoods within the first 7years after PhD graduation, with gender differences increasing over the observation period. A decomposition indicates that variations in publishing, parenthood and parental leave account for ∼ 40 per cent of the gender gap in career advancement, suggesting that other factors, including recruitment disparities, asymmetries in social capital and experiences of unequal treatment at work, may also constrain women's careers.

Leaving home across the recent cohorts in Italy: does economic vulnerability due to labour market status matter?

We focus on the process of leaving home among young Italians and aim to discern: (a) the extent to which the potential economic vulnerability associated with certain individual labour market conditions affects the pathway of leaving the parental home; (b) whether the impact of such conditions has changed over generations of young adults. Data come from the pooling of two cross-sectional rounds of the ‘Families and Social Subjects’ survey conducted by the Italian National Institute of Statistics in 2009 and 2016. Specifically, we consider young adults who at the time of interview were between the ages of 20 and 44, for a total of 22,218 respondents. We employ discrete-time event history models, including a competing risks approach. Our findings suggest that labour market status differs in its influence according to reason for exit and gender. While we observe a clearly negative impact of being out of the labour market for both union and non-union related reasons for men, this negative impact is observed only for non-union-related motives for women. Meanwhile, economic uncertainty connected with precarious jobs does not necessarily have a negative impact on the risk of leaving parental home, both for union and non-union related reasons. Furthermore, we see signs of differences across cohorts relative to this latter set of motives for men. The results highlight the importance of considering reasons for exit in analyses of departure from the parental home.

Partnership Dynamics of LGB People and Heterosexuals: Patterns of First Partnership Formation and First Cohabitation

In this study we examine partnership dynamics among people with different sexual orientations in Germany. More specifically, we explore the process of first partnership formation and first cohabitation among men and women who self-identify as heterosexual, homosexual or bisexual. Given the various discriminations against same-sex lifestyles, and the limited opportunities to meet potential partners, we assume that lesbian, gay and bisexual (LGB) people form partnerships later in life and less frequently than heterosexuals. We further expect that the constantly improving social and legal climate for sexual minorities will lead to a reduction in differences in partnership behaviour by sexual orientation. We use retrospectively reported partnership biographies from the German Socio-Economic Panel, which was supplemented in 2019 with a boost sample of sexual and gender minority households. Using discrete-time event history models, we analyse nearly 15,000 episodes of being single and nearly 20,000 episodes of living without a partner in the household. Around 4.5% of these episodes are from people who self-identify as LGB. The results clearly show that patterns of partnership and coresidential union formation differ by sexual orientation. People with a homosexual orientation—and to a lesser extent people with a bisexual orientation—are less likely to enter into a first partnership and a first cohabitation than people with a heterosexual orientation. Significant changes occur across cohorts: LGB people from younger birth cohorts enter (cohabiting) partnerships much earlier and more frequently than those from older cohorts. Thus, the union formation patterns of LGB and straight people have converged slightly.

Gender Differences in the Protective Role of Grandparenting in Dementia Risk.

This study provides one of the first national longitudinal studies of the association between caring for grandchildren (i.e., grandparenting) and the risk of dementia in the U.S., with a focus on gender-specific variations. We estimated discrete-time event history models, drawing upon data from the Health and Retirement Study (2000-2016). The analytic sample included 10,217 community-dwelling White and Black grandparents aged 52 years and older at baseline. Noncoresident grandparenting was associated with a lower risk of dementia for both women and men compared to grandparents who did not take care of grandchildren. However, the cognitive advantage showed different patterns based on gender and the combination of care intensity and family structure. Grandmothers had a lower risk of dementia than noncaregiving grandmothers when providing a light level of noncoresident grandparenting, while grandfathers who provided intensive noncoresident grandparenting had a reduced risk of dementia compared to their noncaregiving counterparts. Grandparenting experiences within multigenerational households and skipped-generation households were not associated with dementia risk. Intergenerational caregiving plays a pivotal role in shaping cognitive health during later life; however, the impact is nuanced, depending on factors such as gender, care intensity and family structure.

Comprehensive vaping bans are associated with lower odds of initiation into electronic nicotine delivery systems use among young people.

E-cigarette and other electronic nicotine delivery systems (ENDS) use has grown considerably over the past decade, with notable increases among young people. US state policy contexts for ENDS and tobacco may shape initiation into ENDS use among adolescents as they age into early adulthood. We aimed to determine whether state-level comprehensive vaping ban policies reduce the odds of youth initiation into ENDS use, net of additional state-level ENDS and tobacco policies, as well as the youth's cigarette smoking status. Longitudinal data from the Population Assessment of Tobacco and Health study were merged with a state-year database on tobacco and ENDS policies. Multivariable discrete-time event history models of ENDS initiation were estimated. United States. Twenty thousand twelve youth assessed over six waves from 2013 to 2019 (n = 53 974 observations). We examined comprehensive indoor vaping bans (i.e. 100% vape-free workplaces, restaurants and bars) as a key factor in initiation into ENDS use (i.e. first instance of vaping) from age 13 to 22. Among young people, residing in a state with a comprehensive vaping ban was associated with 18% lower odds of ENDS initiation (odds ratio = 0.82; 95% confidence interval =[0.71, 0.94]), even after controlling for other state ENDS and tobacco policies, the youth's cigarette smoking and socio-demographic background and state-level covariates. In the United States, state-level vaping bans (i.e. 100% vape-free workplaces, restaurants and bars) are associated with reduced odds of youth initiation into electronic nicotine delivery systems use.

Leaving the parental home during the COVID-19 pandemic: the case of Southern Europe

In 2020, COVID-19-related governmental restrictions forced individuals to radically change their habits, possibly impacting on their living arrangements. Whether COVID-19 affected young adults’ propensity to leave the parental home is still unknown; Southern Europe is of particular interest, as youth experience the “latest-late” transition to adulthood, face uncertainty in the labor market, and receive low welfare support. Using EU-SILC longitudinal data from Greece, Spain, Italy, and Portugal, this study examines how home-leaving rates evolved in the short-term and explores the relationship between governmental restrictions, economic characteristics of households and young adults, and leaving home behaviors. Descriptive analyses reveal that the share of young adults leaving the parental home in Southern Europe between 2019 and 2020 slightly increased compared to previous years. Discrete-time event history models show that the propensity to leave the parental home increases with the stringency of policy measures. Young adults with the highest likelihood to leave home are employed individuals whose households are in the lowest income quintile as well as students from the highest income quintile, suggesting that, in these countries, residential independence is associated with either the acquisition of economic resources in the labor market or the availability of family resources. We interpret this result in favor of an “independence effect” exerted by COVID-19-related restrictions on young adults; future research might establish whether this trend is temporary or persistent over time.

Education, health indicators and fertility outcomes: a longitudinal analysis of couples in Britain.

Previous studies have shown that highly educated women are more likely to realise their fertility aspirations, or experience a faster progression to a higher order birth, compared to lower educated women. This is often explained by improved economic or social resources among the higher educated. However, it is unclear whether educational differences in health behaviours may also contribute to these differential fertility outcomes. In this study, we use data from Waves 1-7 of the UK Longitudinal Household Study, combined with data from the Nurse Health Assessment from Wave 2 to estimate couples' likelihood of experiencing additional childbirth within six years. A discrete-time event history model is employed to analyse the transition to a higher order birth, while accounting for both partners' level of education as well as smoking patterns and body mass index. We find that couples in which the female partner is highly educated are more likely to experience childbirth within six years compared to others. In addition, female smoking is negatively associated with the likelihood of childbirth, while no significant effect has been found for male health factors. Female health indicators explain some of the variation in fertility outcomes for women with lower secondary education compared to degree-educated women. However, education remains a significant predictor of the transition to higher order births, also after accounting for male and female health indicators. It is therefore important to consider both socio-economic and health factors in order to understand variations in fertility outcomes.

Who is aging out of place? The role of migrant selectivity in international retirement migration

ABSTRACT International retirement migration gained popularity with the rise of globalisation and individualisation, but little is known about whom the retirement migrants are compared to retirees who do not migrate. To gain insight into who migrates compared to who stays, we examine a broad set of individual determinants. We collected data for the survey of Dutch Retirement Migrants Abroad, a new dataset based on a probability sample of Dutch nationals with an oversample of retirement migrants (ages 66–90). The survey includes 5225 migrants who migrated from the Netherlands and permanently reside in one of forty different destination countries and 1339 Dutch retirees who reside in the Netherlands. Using discrete-time event-history models, we test the effect of socioeconomic status, social ties, personality traits, and cultural values on the likelihood of migration. Having a partner and a higher occupational status raised the likelihood of migration. Additionally, retirement migrants were more likely to be adventurous, postmaterialist, and identify with counterculture of the sixties, such as being involved in the hippie culture, than non-migrants. Having more social ties in the Netherlands decreased the likelihood of migration. This study highlights the complex interplay of determinants influencing who migrates at older ages and who stays.

The changing social gradient of marriage and cohabitation in seven Latin American countries

While research shows that cohabitation has increased significantly among highly educated individuals in Latin America, much less is known about how the relationship between educational attainment and first union formation has changed over time and across the region's countries. Accordingly, this paper describes the changes across cohorts in the type of first union (marriage or cohabitation) entered by women from seven Latin American countries. It also analyzes trends in the relationship between women's education and the type of first union within and between these countries. Using Demographic and Health Survey (DHS) data, life tables, discrete-time event history models, and predicted probabilities were estimated to analyze the changing determinants of first-union formation. The results pointed to an overall increase in first-union cohabitation over time, with some important differences across countries. The multivariate analysis suggested that women's education influences the type and sequencing of the first union, with socioeconomically disadvantaged women increasingly likely to transition to early unions and enter cohabitation rather than marriage.

A Diffusion Network Event History Estimator

Research on the diffusion of political decisions across jurisdictions typically accounts for units’ influence over each other with (1) observable measures or (2) by inferring latent network ties from past decisions. The former approach assumes that interdependence is static and perfectly captured by the data. The latter mitigates these issues but requires analytical tools that are separate from the main empirical methods for studying diffusion. As a solution, we introduce network event history analysis (NEHA), which incorporates latent network inference into conventional discrete-time event history models. We demonstrate NEHA’s unique methodological and substantive benefits in applications to policy adoption in the American states. Researchers can analyze the ties and structure of inferred networks to refine model specifications, evaluate diffusion mechanisms, or test new or existing hypotheses. By capturing targeted relationships unexplained by standard covariates, NEHA can improve models, facilitate richer theoretical development, and permit novel analyses of the diffusion process.

Spouses' division of labor and marital stability: Applying the multiple-equilibrium theory to cohort trends of divorce in East and West Germany

Objective: In comparing East and West Germany, we investigate task specialization and its association with marital stability twofold: (1) Has the association between women’s employment and divorce risk changed across marriage cohorts? (2) Are men’s levels of engagement in domestic tasks associated with divorce risk? Background: While older theories assumed that women’s employment destabilized marriages, newer theories suggest that men can re-stabilize marriages by changing their behavior and engaging in housework. Method: We analyze data from the SOEP using discrete-time event history models in a historical and a dyadic perspective. Results: Our results show that the associations between women's employment and the risk of divorce have been changing across marriage cohorts, and that this trend began earlier in East Germany. Husbands' relative contribution to division of housework is not found to stabilize marriages in East and West Germany, but we find differences between marriage cohorts in West Germany. Conclusion: Our findings confirm that the traditional male breadwinner model is no longer associated with a stable equilibrium in marriage in Germany. It appears that either the German society is still in the transitional stage, as men’s contributions to housework are shown to be irrelevant for marital stability; or that gender equality is not associated with the new stable equilibrium in marriages.

Repartnering of women in the United States: The interplay between motherhood and socio-economic status

We examine the socio-economic differentials in mothers’ and non-mothers’ repartnering behaviours following the dissolution of a co-residential (marital or cohabiting) union. Based on five waves of the National Survey of Family Growth (N = 11,479), we use discrete-time event history models, jointly modelling exit from a partnership and entry into a new union. Few differences are found for entry into direct marriage, which is a rarely observed event. However, when we examine women’s entry into cohabitation (a possible stepping stone to marriage), we observe: (1) a motherhood gap, where mothers are less likely to repartner than non-mothers; (2) a negative association between educational attainment and repartnering probability; and (3) the motherhood gap existing only for low-educated women. Supplementary analyses on the impact of the Great Recession demonstrate that whereas the economic cycle mattered for the repartnering of low-educated women, it made no difference for more highly educated women.

Sibling estrangement in adulthood

Whereas interest in adult sibling relationships has been growing, we are not aware of any quantitative studies focusing on sibling estrangement (that is, lack of contact or emotional closeness). This gap in the literature seems unfortunate, because estrangement in close family relationships has been suggested to be momentous for family functioning and individuals’ well-being. Drawing on four waves of the German Family Panel (pairfam; n = 5,729), covering a 6-year observation period, we therefore assess respondents’ estrangement from up to four siblings in adulthood, focusing on the predictive role of core structural parameters of sibling dyads (especially genetic relatedness) and disruptive family events (particularly parental separation/divorce or death). Whereas 28% of respondents in our sample experienced at least one episode of estrangement from any sibling, estrangement occurred multiple times in only 14% of sibling dyads. Moreover, results derived from discrete-time event history models indicate (a) that genetic relatedness is the single most important risk factor, albeit strongly mediated by childhood co-residence, and (b) that sibling relations tend to become more “vulnerable” over the life course, when adverse family events accumulate.

Inequalities in Retirement Life Span in the United States.

The length of retirement life may be highly unequal due to persistent and significant discrepancies in old-age mortality. This study assesses gender and educational differences in the average retirement life span and the variation in retirement life span, taking into account individual labor force exit and reentry dynamics. We used longitudinal data from the Health and Retirement Study from 1996 to 2016, focusing on respondents aged 50 and older (N=32,228). Multistate life tables were estimated using discrete-time event history models. The average retirement life span, as well as absolute and relative variation in retirement life span, were calculated analytically. Among women, we found a persistent educational gradient in average retirement life span over the whole period studied; among men, the relationship between education and retirement expectancy differed across periods. Women and the lower-educated had higher absolute variation in retirement life span than men and the higher-educated-yet these relationships were reversed when examined by relative variation. Our multistate approach provides an accurate and comprehensive picture of the retirement life span of older Americans over the past two decades. Such findings should be considered in high-level discussions on Social Security. Potential reforms such as raising the eligibility age or cutting benefits may have unexpected implications for different social groups due to their differential effects on retirement initiation and reentry dynamics.

Cold homes and mental health harm: Evidence from the UK Household Longitudinal Study

Cold homes are associated with a range of serious health conditions as well as excess winter mortality. Despite a comparatively mild climate cold homes are a significant problem in the UK, with a recent estimate finding that over one-quarter of low-income households had been unable to adequately heat their home in winter 2022. The magnitude of cold housing in a country that benefits from a mild climate indicates indifference towards, or acceptance of, a significant minority of people living in inadequate conditions on the part of policy makers. Cold homes are therefore a source of social harm. Recent changes to the household energy price cap, the rising cost of living, the ongoing effects of the benefit cap, and below inflation uprating to social security benefits is likely to greatly exacerbate this issue. In this research we use data from the UK Household Longitudinal Study to explore whether living in a cold home causes mental health harm. We control for mental distress and housing temperature on entry to the survey in order to account for the potentially bi-directional relationship. Multilevel discrete-time event history models show that the transition into living in a home that is not suitably warm is associated with nearly double the odds of experiencing severe mental distress for those who had no mental distress at the beginning of the survey; and over three times the odds of severe mental distress for those previously on the borderline of severe mental distress. These results show the significant costs of failing to ensure that people are able to live in homes in which they are able to live comfortably by even the most basic standards. These costs will be felt not just individually, but also more broadly in terms of increased health spending and reduced working.

Does local income and wealth inequality affect mortality? A register-based fixed effects study of 58 million person-years.

Aim: The Income Inequality Hypothesis asserts that income inequality causes negative health burdens within wealthy nations. We aimed to test the hypothesis in Norway, examining the relationship between both income and wealth inequality, net of individual economic resources, and individual all-cause mortality. To this day, little is known about the association between wealth inequality and mortality. In Norway, wealth is far more unequally distributed than income and can be a more prevalent indicator of long-term financial security. Methods: We estimated discrete-time event history models from Norwegian register data, covering all Norwegian-born men and women aged 25-84 years during the period 1993-2013. We include fixed effects at the municipal level, adjusting for time-invariant characteristics in the local area. Our data contain approximately 58.5 million person-year observations after sample restrictions. Results: Overall, increased income and wealth inequality (measured using Gini coefficients of individuals' pensionable income and net worth) is not associated with higher mortality risk in Norway. With a one percentage point increase in income inequality, odds ratio (OR) for mortality was estimated at 0.998 (95% confidence interval (CI) 0.996-1.001) for men and 1.002 (95% CI 0.999-1.006) for women. Corresponding OR from increased wealth inequality was 0.998 (95% CI 0.997-1.000) for men and 0.999 (95% CI 0.998-1.001) for women. Conclusions: Our results raise further doubts about the existence of a contextual effect of economic inequality on health, at least in a Nordic setting. Extensive health policies, relatively generous benefits and high levels of trust could potentially function as a buffer against negative health consequences of increased economic inequality in Norway.

Variation in the link between income position and union dissolution: an analysis of couples with and without a migrant background in Belgium

ABSTRACT Despite the fact that the size and diversity of migrant populations and their descendants in Europe have increased substantially throughout the past decades, income dynamics of union dissolution have not been studied among populations with a migrant background in the European context. Using data from the Belgian Social Security Registers from 1999–2013, we estimate discrete-time event-history models of union dissolution for couples of Belgian, Southern European, Moroccan and Turkish origin. The results indicate a weaker link between household income and union dissolution among couples with a Southern European and Turkish background. Whereas couples without a migrant background experience the lowest union dissolution risks when incomes are evenly distributed, the male breadwinner model does not raise union dissolution risks of couples with a migrant background. Welfare dependency has a positive association with union dissolution among all couples but the threshold at which it raises union dissolution risks is higher for couples with a Moroccan or Turkish background.

The Influence of Parental Education on First Marriages in China: The Role of Childhood Family Background

China has implemented the universal two-child policy to promote fertility whereas the trends of later marriage and childbearing still prevail. The marriage timing of young adults has received considerable attention. However, studies tailored to the Chinese context remain limited. Using data from the China Labor-force Dynamic Survey 2016, we established bivariate and multinomial logistic regression discrete-time event-history models to examine the influence of parental educational attainment on their children’s first marriages. Not only the effects of parental education on their children’s likelihood of entry to first marriages and marriage postponement but also changes in these effects by sex, region, and across birth cohorts and the individual life course were analyzed. In general, there was a significantly positive relationship between parents’ educational attainment and the likelihood of their children’s entry into first marriages. However, it couldn’t explain variations about the timing of first marriages or marriage postponement. Analyses of subsamples revealed considerable heterogeneity in the effects of parental educational attainment by sex and region. These findings reveal the influence of childhood family background factors on marriage behaviors and provide a basis for predicting future marriage and childbearing trends in China. This study provides meaningful inputs and a rationale for amending the minimum legal marriageable age and for promoting marriage and births within Chinese marriage law.

Is there a welfare trap? Duration dependence in social assistance recipiency among young Finns

The probability of social assistance exit decreasing over time is called negative duration dependence. This is one of the most debated aspects of social assistance receipt. This study analyses duration dependence among young adults in Finland and aims to show how social assistance receipt per se affects the likelihood of exiting social assistance. It also examines whether parental recipiency is associated with duration dependence. While the intergenerational transmission of social assistance has received extensive research attention, there is a scarcity of evidence on whether parental recipiency is associated with social assistance exits and duration effects. This study uses full monthly history of social assistance receipt between the ages of 19 and 29 years. The analyses are based on Finnish register data and conducted using discrete-time event-history models (pooled logistic, random-effect logistic and fixed-effect logistic). The findings indicate that controlling for all time-invariant characteristics significantly reduces but does not completely remove duration dependence among young adults. They also show that those whose parents received social assistance are less likely to exit social assistance. In models controlling for all time-invariant characteristics, parental social assistance receipt is also linked with stronger duration dependence. The implications of these findings are discussed.