The impact of economic uncertainty, precarious employment, and risk attitudes on the transition to parenthood

Schmitt, Christian Article — Accepted Manuscript (Postprint) The impact of economic uncertainty, precarious employment, and risk attitudes on the transition to parenthood Advances in Life Course Research Provided in Cooperation with: German Institute for Economic Research (DIW Berlin) Suggested Citation: Schmitt, Christian (2021) : The impact of economic uncertainty, precarious employment, and risk attitudes on the transition to parenthood, Advances in Life Course Research, ISSN 1040-2608, Elsevier, Amsterdam, Vol. 47, pp. --, https://doi.org/10.1016/j.alcr.2021.100402 This Version is available at: https://hdl.handle.net/10419/270939 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ The Impact of Economic Uncertainty, Precarious Employment, and Risk Attitudes on the Transition to Parenthood Christian Schmitt German Institute for Economic Research (DIW Berlin) Rostock University & City of Leipzig (word count: 7984 excluding tables, figures, and references) Abstract: This study investigates how precarious employment throughout the life course affects the fertility behavior of German men and women, and how risk attitudes moderate exposure to objectively given uncertainty. Analyzing data from the German Socio-Economic Panel Study (SOEP) from 1990 to 2015, I find that men and women have become quite similar in their fertility behavior: stable employment accelerates family formation, whereas discontinuous and precarious employment delays it. With regard to risk attitudes, risk-averse women show the highest likelihood of family formation. Apparently they choose a family and parenthood centered pathway in their life course that provides them with stability and social approval, where an uncertain and instable employment career cannot provide these resources. Keywords: economic uncertainty, precarious employment, risk attitudes, Germany, parenthood decision, life-course, event-history analysis. This is the postprint of an article published in: Advances in Life Course Research 47 (2021), 100402, 14 S., available online at: https://doi.org/10.1016/j.alcr.2021.100402 © <2021>. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/ Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 2 Introduction This study investigates the impact of economic uncertainty and precarious employment in male and female life courses on first-parenthood decisions. The period of analysis is 19902015. This includes the political transformation in the early phase of German reunification, as well as the labor market reforms in the early 2000s that resulted in a liberalization of key labor market institutions. The goal of this study is to delineate the underlying mechanisms shaping fertility within the scope of related uncertainties. More specifically, I aim to unravel how labor market-related uncertainty and precarity in individual employment biographies affect individual fertility behavior in the life course. This study aims to extend the literature on the employment-fertility nexus by honing in on the role of individual risk attitudes. In particular, the specific contribution of this study is to consider individual risk attitudes as possible moderators in the relation between employment uncertainties and fertility decisions, where previous studies assumed such uncertainties to affect all individuals in a similar way. The theoretical foundation follows an approach of life-course decision-making that relates the pathway of family formation to preceding stages and future prospects of work biographies. Taking into account the multidimensionality of the life course (Buhr & Huinink, 2014), I consider the accumulation of disadvantageous conditions in individual employment biographies and corresponding life course choices with regard to family formation. In this vein, I will investigate how the personality trait of risk aversion moderates different forms of uncertainty that might affect childbearing decisions in the life course. The empirical analyses rely on data from the German Socio-Economic Panel (SOEP) study, a longitudinal survey capable of capturing extensive parts of individuals’ fertile 3 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany lifespan over the course of more than 25 years. Applying an event-history approach, the data make it possible to investigate the role of several key factors related to fertility choices, including personality traits, social and economic status, precarious employment contexts, and partnership characteristics. This article is structured as follows. The subsequent analytical framework elaborates the micro-theoretical perspective on a decision-making framework of family formation within male and female life courses. This is extended with a discussion on how economic uncertainties and precarious employment affect fertility behavior, and a review of the research literature. In the latter context, I also examine how risk attitudes might moderate the uncertainty-fertility nexus. The subsequent discussion of the institutional background and policy context provides information on work-family cultures in Germany, on East-West German differences in culture and institutions, and on changes in welfare state and labor market institutions that are capable of influencing fertility behavior. The database and design of the empirical analysis are outlined in the data and methods chapter, followed by a presentation and discussion of findings. Analytical framework A life-course-related decision-making framework In this study, I focus on parenthood decisions, that is, the transition from childlessness to parenthood. The analytical framework of this study follows the assumption that childbearing decisions are result of rational choice (Leibenstein, 1981, Lindenberg, 1985). One could Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 4 challenge the notion that childbirth is the consequence of a decision-making process. However, even future parents confronted with an unplanned pregnancy have to make a decision at some point.1 In general, the transition to parenthood in western societies is a consequential and irreversible choice in the life course, usually accompanied by an intensive consideration of various factors (Hobcraft & Kiernan, 1995; Hagestad & Call, 2007; Huinink & Kohli, 2014). This includes considering the traits of a suitable partner with whom to start a family and aligning competing biographical plans and options (Mynarska, Matysiak, Rybińska, Tocchioni, & Vignoli, 2015). One could argue that the same applies to decisions revolving around having second or further children. However, first-parenthood decisions crystallize the conflicts involved in reconciling competing life goals, such as career and family paths. Fertility decisions and precarious employment When starting a family, gainful employment mainly functions in preparing the economic foundations. Contributions to family income are increasingly provided by both men and women. But while women are gaining ground in the labor market, the exclusiveness of the female caregiver role is rarely challenged. Women in Germany face strong “good mother” norms (Trappe, Pollmann-Schult, & Schmitt, 2015). As a result, the opportunity cost of foregone income due to childcare leads to a motherhood of parenthood among high-income e. In this study, I investigate labor market precarity in relation to different forms of insecure or uncertain work contexts, including unemployment, working reduced hours or 1 One might argue that such a delayed choice in the case of an unplanned pregnancy involves a host of ideational and moral aspects that remain hidden in the survey-based investigation of fertility. In this case, the estimates would be biased by an increased number of unobservables associated with such considerations. 5 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany on an irregular schedule at or below the poverty level (marginal employment), as well as other work contexts that display a weakened link to the labor market, such as temporary work (Kalleberg, 2009). The threat emanating from such insecurities has been increasing with the reduction in welfare state’s efforts to hedge market risks (Breen, 1997). Among men, the increase in insecure and precarious work (Grotti & Scherer, 2014) has significantly undermined their previously undisputed breadwinner status. It should be noted that – for the sake of this study – the norm of permanent stable full-time employment serves as a frame of reference. Importantly, both standard and nonstandard work can be precarious. However, non-standard work is more often uncertain as it is frequently associated with bad working conditions and low wages (ILO, 2016). Regardless of precariousness, non-standard work frequently provides earnings that are insufficient to take the role of a family breadwinner2. The question of how precarious work contexts affect the likelihood to become a parent has spawned a large body of literature. Much of the research goes beyond demographic issues to address the question of how labor-market-induced uncertainty in postindustrial societies affects people’s willingness and readiness to make lasting commitments. The standard employment pattern among men implies a norm of permanent fulltime work as prerequisite for supporting a future family. In this context, working on fixedterm or short-term contracts and periods of unemployment or labor market inactivity are detrimental to a person’s socio-economic status (Green, 2008; Giesecke, 2009) and 2 In order to establish distinct groups, the empirical concepts used in this study distinguish permanent fulltime employment as standard work, from non-standard or atypical work. The latter group summarizes parttime work, and precarious work (including fixed-term jobs regardless of work hours, and marginal work). For a more detailed definition of groups cf. p. 18f. Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 6 threaten their future ability to make investments in children’s needs. In such situations, characterized by prevailing uncertainties, childbearing tends to be postponed (Scherer, 2009; Kreyenfeld, 2010). The social norm of stable full-time employment, however, is less rigid for women. Being in an atypical employment context (such as part-time or marginal work) might tip the scales in favor of motherhood, particularly when this type of employment offers more flexibility in time budgets at a lower price of time. This is the case, for instance, with part-time work or unemployment. Additionally, the social approval gained from parenthood might compensate for the status loss associated with precarious employment or joblessness. But in many countries, women face a higher risk of precarious employment, which makes them more vulnerable to social and economic exclusion in highly commodified welfare systems (Barbieri, 2009). Hence, starting a family from a precarious or unstable employment position remains an uncertain endeavor, even if family formation plans predominate. Key research findings on the fertility-employment nexus presented in Table 1 appear inconsistent at first glance. However, much of the variation relates to different mechanisms associated with different types of precarious employment. For example, fixed-term contracts may have little impact on current income but may limit the predictability of future income. Unemployment, in contrast, significantly reduces family income, but increases the amount of time available. This observation sheds light on the puzzlement of differing findings across countries: The fertility effects of precarious employment crucially depend on the generosity and universalism of the respective welfare system, employment protection regulations, and work-family cultures. Unemployed women in France have been shown to 7 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany postpone parenthood as the French welfare state provides comprehensive childcare services and promotes female labor market attachment, (e.g., Pailhé & Solaz, 2012b). For German women, in contrast, most studies find an increased likelihood to start a family after job loss. This is likely due to a generous unemployment support system and a low coverage and acceptance of public childcare (Özcan, Mayer, & Luedicke, 2010; Schmitt, 2012b; Kreyenfeld & Andersson, 2014; Schneider, 2016). A further variation in research findings regards the study of first or higher-order births (Table 1). This limits comparability since family formation follows different rationales as family extension. Importantly, differences across the research literature are also driven by inconsistent measures of employment status. Many studies fail to state which labor market groups are pooled together in the group of economically inactive, unemployed, or non-employed individuals. Moreover, results will differ if analysis of the employment status uses either full-time employment, non-employment, or working hours as a reference category. All these variations appear in the research literature referenced in Table 1. Hence, what is interpreted as a gender effect, a welfare state effect, or a labor market effect, is often related to methodological choices in the studies compared. As a consequence, it would be advisable for future research to interpret findings using standard (full-time) work as reference when compatible with the research focus. Within this general framework, the level of educational attainment moderates the role of job insecurities by moderating whether acquiring a stable job takes priority over starting a family. Placing family formation in phases of precarious employment bears the risk of consolidating labor market exclusion and the devaluation of human capital. Among the few studies that have investigated interactions between unemployment and starting Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 8 a family, results for men do not show a clear pattern across countries (Kreyenfeld, Andersson, & Pailhé, 2012; Schmitt, 2012a; Schmitt, 2012b; Kreyenfeld & Andersson, 2014). For women, some of the findings are either insignificant (Pailhé & Solaz, 2012a, and Schmitt, 2012a for France; de Lange, Wolbers, Gesthuizen, & Ultee, 2014 for the Netherlands), or inconclusive (Özcan, et al., 2010). In the countries for which positive effects of unemployment on first motherhood have been reported, a clear pattern emerges in Denmark, Germany, the UK, and Finland, separating tertiary education (neutral or negative) from medium and lower education (neutral or positive; Schmitt, 2012a; Schmitt, 2012b; Kreyenfeld & Andersson, 2014; Huttunen & Kellokumpu, 2016). Table 1: Fertility effects of uncertain and atypical employment contexts Study Country/Data Focus Fixed-term contract Unemployed Working part-time Inactivity/ Other M W M W M W M W Liefbroer & Corijn, 1999 Netherlands Flanders Young adults o - o o Hoem, 2000 Sweden - REG o Andersson, 2000 Sweden - REG Age 20-30 + Hank, 2001 Sweden - REG First birth Higher p arities + o Santow & Bracher, 2001 Sweden - FFS First birth o Kravdal, 2002 Norway - REG First birth Higher parities - + + - Meron & Widmer, 2002 France - YCS First birth - - + Baizán, Aassve, & Billari, 2003 Spain - FFS First birth + Tölke & Diewald, 2003 Germany - FS First birth o - - Self empl . - Kreyenfeld, 2004 Germany - FFS First birth + East West - + Vikat, 2004 Finland, age<31 - REG age>30 First birth + - - - Kurz, Steinhage, & Golsch, 2005 Germany - SOEP First birth o o - o o + - o Mencarini & Tanturri, 2006 Italy Multinomial logit of parity - - 15 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany male earners’ low or unstable income in the context of precarious and discontinuous employment patterns. However, in combination with spousal tax splitting and subsidiarity on the household level, many of the changes fostered the emergence of jobs with limited working hours and pay (mini and midi jobs), which couples often used to supplement the income of a full-time earner. The consequence was an increase in marginal and part-time work among women who had previously held the role of full-time homemaker (Gottschall & Schröder, 2013). Agenda 2010 also embedded the adult worker model in German social policy as a means of implementing one of the 2001 Lisbon Strategy goals. According to this model, men and women were to be treated equally as part of the workforce, subject to all the same obligations and sanctions in the case of benefit dependency. However, the labor market reforms neglected to provide institutional support enabling parents to fulfill the care responsibilities for which women have traditionally been responsible. This gendered division of labor has been reproduced for decades particularly in West Germany. Under this scheme, women were encouraged to retreat from the labor market or work part-time by seemingly generous leave policies, a spousal tax-splitting that (still) favors a singleearner model, and a limited childcare supply. This perspective has – in part – been changed since 2007 with a series of family policy reforms. These policies mark a shift towards increased gender equity and the promotion of work-family balance closely oriented toward Scandinavian policies. (Spiess & Wrohlich, 2008). The 2007 parental leave benefit reform (Elterngeld) encourages paternal contributions to childcare (through “daddy Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 16 months”) and a net income replacement of two-thirds with a comparatively short payment period (12 months). This encourages a swift return to the labor market4. This policy reform has been flanked by an expansion of public childcare services since 2009, and a legal entitlement to a daycare slot for children under the age of three since 2014 (Schober & Schmitt, 2017). Data and Methods The analyses use data from the German Socio-Economic Panel (SOEP) study (Wagner et al., 2007). Since 1984, this nationwide household panel has surveyed between 6,000 and 18,000 households across Germany each year. The SOEP data enable analysis of various objective and subjective factors affecting fertility. They include household composition, employment histories, health, education, and fertility histories. The empirical models are based on a continuous time event-history analysis of the transition from childlessness to parenthood, that is, the decision to become a parent. The population at risk consists of childless men and women in the 1965-1989 birth cohorts observed during the period 1990-2015. This choice essentially encompasses the respondents who were born after the baby boom and before German reunification5. The analysis is restricted to the 16-45 age range, since transitions to parenthood above and below this age span are rare exceptions in Germany. The selection criteria result in a sample of about 10,000 men and women with about 2,800 parenthood decisions observed during the time 4 Previous studies have not found comprehensive evidence of the leave reforms resulting in changes in fertility behavior to date. The exception are short-term bandwagon effects around the time of implementation (Tamm, 2013; Cygan-Rehm, 2016). 5 We focus our sample selection on the representative core-samples A-F & H (Germans, foreigners, East Germans, migrants, and refresher samples 1998, 2000 & 2006), excluding the snowball-sampled population within the migrant sample D. 17 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany of analysis. The SOEP data comprise both annual and monthly data (chronological age, age at childbirth, employment status, changes in partnership status). Each month during the observation period depicts a temporal unit of analysis. Annual data is assigned to each monthly variable on the basis of the month of interview closest to a monthly record. Key variables The dependent variable is the timing of the decision to have a first child. This time-dependent process is modeled using an event-history approach. The SOEP provides detailed fertility histories of men and women, including the year and month of birth of both the parents and child. We approximated the time of the decision to have a first child by backdating 10 months6 from a respondent’s age at the time of childbirth. Since parenthood decisions are restricted to observed births, there is a potential bias against parenthood decisions that do not end in delivery. This may lead to slightly inflated standard errors. The time-constant covariates include a period control with four-year-segments, which aims to capture macro-level heterogeneity, and specifically political and economic changes as institutional shifts are closely linked to life-course outcomes (Mayer, 2009), the transition parenthood in particular. Moreover, I consider a basic cohort distinction with the goal of controlling for changes in attitudes and socialization (e.g., regarding care) across birth cohort. This indicator, however, only distinguishes between older and younger cohorts (1965-1975 and 1976-1989) to avoid an overspecification bias. It should also be noted that age, period, and cohort aspects still intersect with a persistent identification problem 6 It is key in this context to consider a covariance matrix that represents the individual status preceding (and not following) the point of parenthood decision. Schmitt (2012a) conducts sensitivity tests that yield robust results when backdating between 9-12 months. This procedure is also in line with medical evidence showing that about 70% of couples are able to conceive within 1-3 cycles (see, e.g., Gnoth, Godehardt, Godehardt, Frank-Herrmann, & Freundl, 2003). Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 18 in linear dependency of these factors (Yang & Land, 2008). The analyses also consider migration background7 and whether a person grew up in East or West Germany8 (thus covering different cultural imprints with respect to (female) employment, and work-family balance). Moreover, I control for growing up with siblings or as an only child, since this might shape perspectives on family life and having children. The time-varying controls include partnership status (single or cohabiting) and educational attainment. The latter is included as a variable with three stages: basic education (i.e., lower secondary), intermediate education (i.e., upper secondary) and higher education (i.e., college degree /other tertiary). The key aspect of this study is the empirical investigation of objectively given (workrelated) hardships and uncertainties and their subjective assessment (on the basis of personal risk attitudes). In this endeavor, I distinguish between past experiences that may also affect future work prospects (e.g., unemployment histories) present labor market statuses that indicate economic or job-related hardships (e.g., marginal work or current unemployment), and present conditions, which indicate future uncertainties (like fixed-term contracts) that might threaten the ability to support a family. In order to differentiate between these aspects, a number of indicators have been integrated into the model estimates. The 7 This combines first and second-generation migrants, but due to the combination of the age ranges considered in this study and prime years of collecting migrant samples in the SOEP (1984 & 1994), second-generation migrants constitute the majority (SOEP_Group, 2018). 8 Variable: loc1989 (SOEP_Group, 2018). 19 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany employment status indicator differentiates full-time employment with permanent contract (reference), full-time with a temporary or fixed-term contract9, part-time employment, marginal employment (working fewer than ten hours per week or on an irregular basis), as well as registered unemployment, economic inactivity (including homemakers), and attending school or university. The differentiation between full-time work with a permanent contract (as reference category) and full-time employment with a temporary or fixed-term contract (as one of the additional categories) is specifically introduced to control for looming uncertainties in fixed-term jobs (in particular, potential job loss), which are not necessarily associated with other occupational disadvantages such as reduced work hours or low pay. Furthermore, fixed-term employment is considered only for full-time jobs in order to highlight this element of uncertainty in what otherwise constitutes a standard employment relationship. While part-time work (particularly among men) and marginal work are frequently associated with fixed-term contracts, I argue that precarity already dominates these forms of atypical work10. Additional controls associated with earner capability and occupational uncertainty include a couple’s joint available (i.e., net) income in a given month, and a continuous indicator of accumulated unemployment experience (indicating the fraction of time spent 9 Whether a respondent was working on basis of a fixed term employment was collected in the SOEP from 1990-1994 among persons entering a new job or for first-time respondents. When an existing work contract was changed from fixed-term to permanent, this change was not recorded between 1990-1994. As a consequence, there might be an overreporting of fixed-term contracts 1990-1994. This situation has been remedied in 1995 and subsequent waves, by asking any employed individual for permanency of work contract. 10 Other forms of parallel activities are considered in the hierarchy presented in the models (Tables A2-A3). An unemployed person in a training program, for instance, would be considered unemployed; a college student working part-time would be considered a part-time worker according to this status hierarchy. Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 20 in unemployment since age 25 on a continuum from zero to one, where one equals permanent unemployment). As outlined above, I aim to consider how objectively given uncertainty is moderated by individual risk perception. The SOEP surveyed risk attitudes on an 11-point Likert scale in 2004, 2006, and annually since 2008. The question is “How do you see yourself: Are you generally a person who is fully prepared to take risks or do you try to avoid taking risks?” The scale of answers runs from 0 (“unwilling to take risks”) to 10 (“completely willing to take risks”). This abstract phrasing implies limitations in the validity of such a complex and multidimensional characteristic11. However, various studies have confirmed the high construct validity of the item and even recommend an abstract measure, since this way of phrasing questions is not biased by domain-specific risk behavior (Schmidt, 2008; Reynaud & Couture, 2010; Dohmen, et al., 2011)12. As risk attitudes have only been collected in the SOEP since 2004, the question whether risk propensities are stable over the life course is important. Previous research has found that – aside from a gradual increase at a very low rate with age – risk attitudes are a stable characteristic of individual personality (Donkers, 1999; Borghans, et al., 2008; Claudia R Sahm, 2012). More support for this finding is provided by a series of studies on the stability of the Big Five personality traits, with which risk attitudes are closely associated. The findings suggest that the Big Five are stable against most life events including 11 That is, a person may be highly risk-prone in their health behavior and at the same time highly risk-averse in their financial matters (Hanoch, Johnson, & Wilke, 2006). 12 In survey wave 2004, respondents were asked to rate their specific willingness to take risks in their leisure time, when driving a car, in their health behavior, in saving money, in career decisions, and in trusting other people. These items all yield correlations >.8 with the abstract risk item (results available on request). 21 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany childbirth (see Donnellan, Conger, & Burzette, 2007; Cobb-Clark & Schurer, 2012). An investigation by Specht, Egloff, & Schmukle, (2011) did not find significant changes in personality triggered by events addressed in the study at hand, including childbirth and becoming unemployed. The risk indicator used in this study consists of an individual mean value across all survey points and ranges from 0 (highly risk-averse) to 10 (highly risk-prone). To account for age-related changes, individual mean risk is discounted at the rate empirically specified by Dohmen, et al., 201113. Empirical Design The empirical models investigate parenthood decisions among childless individuals using an event-history approach. The time at risk starts at age 16. Figure 2 displays the hazard rate of the first-parenthood decisions across age among childless men and women. In order to minimize a bias to the population at risk due to left-truncation, individuals who enter the survey beyond age 25 are excluded from analysis. The baseline hazard of parenthood decisions has an approximately normal distribution, peaking at around age 28 for women and age 30 for men. We capture this distribution by applying a flexible, semi-parametric Cox proportional hazards model14 (Efron method for ties, Box-Steffensmeier & Jones, 2004) with the following specification: (1) 13 According to Dohmen, et al., 2011, adults become more risk-averse by 0.02 points (on an 11-point scale) with each year of age. If the averaged risk aversion of an individual over a course of 10 years lies at 5.7, the applied discounting would result in a value of 5.6 at the beginning of the 10-year period and 5.8 at the end. 14 I apply a continuous model to discrete data. However, a monthly measurement of an event that occurs once per lifetime is – although intrinsically discrete – approximately continuous. This also applies to the set of time-varying covariates that are unlikely to change on a shorter-than-monthly time frame. 0 ( ) ( )exp( ( ) ( )) j j j ra ra k k h t h t x x t x t       Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 22 where, 𝑥𝛽 refers to the time invariant set of covariates, 𝑥𝛽 refers to the time variant covariates and 𝑥𝛽(𝑡) refers to individual risk tolerance, which is considered to be time varying due to temporal discounting across a the life-course. n of events = 3,168 n of cases = 11,072 Figure 2: Hazards of first-parenthood decisions among childless men and women, cohorts 19601989. Source: SOEP v33, 1990-2015; own estimates. Model design The event history models presented in Tables A1-A3 are all estimated separately for men and women, as well as in a joint model. The theoretical reasoning behind distinguishing between men and women is outlined the analytical framework. Men and women face highly differing social expectations, constraints, and costs when considering the transition to parenthood, particularly in the context of employment choices. Table 1 summarizes re0 .002 .004 .006 .008 Hazard 20 25 30 35 40 Years of age (estimates based on time in months) Women Men 23 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany search findings, many of which distinguish between men and women, with differing effects along gender lines. Recent research (e.g., Kreyenfeld & Anderson, 2015) shows that men and women have become more similar in the linkage between employment and fertility behavior. The joint model, which combines fertility choices of both men and women, is specifically meant to address this. In cases where male and female fertility behavior is similar against a backdrop of employment uncertainties and individual risk attitudes, the joint model should produce similar patterns to the gender-focused models, while being able to rely on greater statistical power due to higher case numbers. The initial models (Table A1) investigate the unconstrained impact of risk attitudes on fertility decisions (Model 1.1) as well as the impact of various controls (Model 1.2). Model 2 (Table A2) additionally investigates the role of employment status. That is, all background controls from Model 1.2. are included, but omitted from display. Model 3 (Table A3) also controls for couples’ joint (net) income to consider the economic resources they have available to support a family (while maintaining the focus on the individual level). Since this indicator is confounded with employment status, it is introduced in a separate model. In addition to these models, interaction effects between risk attitudes and key variables of interest are explored. The results of these models have been visualized in Figures 3-6 (detailed estimates available on request). Results This section presents the key findings from the regression estimates and the results of interaction effects (graphs provided in the addendum). Individuals with a migration background and a low level of education show an increased likelihood to start a family (Model Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 24 1.2). The propensity of first-parenthood decisions is also higher among individuals who grew up with siblings, individuals in the 1965-1975 cohort, and East Germans. The estimates in Tables A2 and A3 specifically investigate the impact of economic uncertainty and different forms of precarious employment. Surprisingly, the results show women having become quite similar to men with regard to employment patterns that suggest a postponement of parenthood due to employment-related uncertainties. The findings suggest that men and women tend to decide to start a family from a position of stable, full-time employment. Still being in education or in atypical or precarious work contexts is linked to a postponement of family formation, with similar patterns being found for men and women. This could be related to a life-course effect, where the completion of education is followed by an establishing-phase in the labor market, and family formation, subsequently. Gender differences in the employment and fertility nexus remain in place in older German cohorts, including those born in the 1950s (Gash & McGinnity, 2007; Kreyenfeld, 2010; Schmitt, 2012b) and have also been reported for several other countries (e.g., Andersson & Scott, 2008 for Sweden). Apparently, many of these gender differences have declined in recent years and among younger cohorts (see also Kreyenfeld & Andersson, 2014). Fixed-term contracts One of the few employment contexts without clear gender convergence is full-time work on a temporary contract. The results are insignificant for men, while women show a 23% reduced rate to start a family compared to full-time working women with a permanent contract. Given the persistent male breadwinner norm in Germany, we would rather have expected a negative link among men rather than among women. However, the findings 31 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany haven” hypothesis. Alternatively, being in atypical work and being highly risk prone might be linked to the accumulation of aspects that inhibit the transition to parenthood (namely incomplete labor market integration and openness for new experiences and sensation seeking as personality traits). A closer focus on individuals, persistently remaining in marginal work over extended periods (Figures 5 & 6) shows that this hampers the realization of family formation plans significantly, and this association is most pronounced among the more risk-averse (i.e., risk attitudes <4 on a 0-10 scale). This is also visualized by a considerably steeper slope, indicating a strong decline in birth propensities in the risk averse group with long duration in marginal employment (Figure 5). Apparently, the discouragement of being trapped in marginal work contexts is perceived as more threatening by women with pronounced risk aversion. Conclusion Marginal employment, an extensive history of unemployment, temporary work, but also part-time work, and low income – basically any employment context that deviates from the template of permanent full-time work – results in a delay of family formation in the life course. Moreover, women have become quite similar to men in their labor market behavior prior to parenthood. The underlying causes are the increase in female educational attainment, but also the need to have a dual-earner model in place prior to family formation due to less stable and predictable employment careers. This unpredictability is also related to a decrease in legal employment protections that emerged during the observation period. Labor market institutions and welfare state policies increasingly subject Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 32 women to the expectations of the adult worker model rather than a female homemaker model. This change in welfare state orientation is reflected in labor market deregulation and more limited social support, particularly since the early 2000s. These developments have given way to a perception of increasing economic uncertainties, and a new family formation model that relies on dual earners. In line with these developments, the findings show a delay in family formation in precarious and otherwise atypical employment contexts. A key issue is the extent to which the German welfare state alleviates the uncertainty associated with unstable employment contexts or work biographies dominated by precarity. The findings suggest that a sequential model of family formation is becoming more prevalent: most family formation occurs from a position of stable, sound employment and only after such a position has been attained. A distinct focus of this study was on the relationship between perceived employment uncertainties and risk attitudes as a personality trait. I find robust results that riskaverse women have a higher likelihood to start a family than their risk-prone peers. One explanation offered in this study is the “family as safe haven” principle (Hypothesis 2), according to which becoming a parent serves as an anchor of stability and a source of social approval under conditions where employment does not provide these resources. Risk averse individuals in particular should pursue this behavior. Personality psychology provides another perspective on this association: Individuals who are more risk-prone show higher levels of openness and extroversion. They might delay family formation in order to avoid limiting social contacts, losing the opportunity for new partnerships, and possibly missing other options beyond family life early in the life course. 33 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany In contrast, risk-prone individuals may be considered to be better suited to endure the strain of economic or societal uncertainty (Hypothesis 1: “family formation as uncertain endeavor", with risk prone individuals being less likely to be constricted by such concerns). They should feel more comfortable with entering the novel and unpredictable life stage of parenthood. The findings – at least for women – clearly point towards Hypothesis 2 (“family as safe haven”): Highly risk averse women show the highest transition rates to parenthood, highly risk prone women the lowest. This picture is exaggerated if atypical work contexts (summary of part-time and marginal employment) are added to the equation. Risk prone women in such jobs show the lowest birth decision rates. This stands in contradiction to the “uncertainty reinforcement” hypothesis (3) that suggest high risk aversion should amplify the perception of uncertainties, thus, reinforcing the delay in family formation. A closer look on the duration in marginal employment exclusively, however, shows that risk attitudes do indeed moderate the role of precarious employment primarily in a context of lasting marginal work. Being trapped in such a highly precarious employment over a longer time reduces the likelihood to opt for parenthood among risk averse women, but not so among risk prone ones. Precarious employment is considered to hamper parenthood transitions as it undermines the ability to provide for a family. The findings show that repeated or excessively long unemployment periods go hand in hand with severe discouragement and weakened attachment to the labor market. The same results are found for individuals who remain in marginal employment over extended periods. Individuals in marginal employment face bleak prospects of not being able to provide for a family, as these irregular Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 34 jobs commonly yield low salaries, while often demanding extensive flexibility in daily routines. Both extended histories of unemployment episodes and extended periods of marginal employment yield a pronounced negative impact on the likelihood to opt for parenthood. In summary this means, while risk proneness leads to a delay in family formation, it also cushions the delaying effect of employment uncertainty in the form extended periods of marginal work. In contrast to previous studies focused on older cohorts, the empirical analyses did not show a positive impact of current unemployment on family formation. Although unemployment hampers earner abilities, a positive association with fertility has been attributed to reduced opportunity costs of childcare during unemployment in a welfare state that limits the ability to reconcile work and caregiver roles (Schmitt, 2012a; Kreyenfeld & Andersson, 2014; Schneider, 2016). A reason for the absence of such a positive relationship might be differing patterns of work-fertility behavior among younger cohorts, which are the focus of this study. However, the analyses show a positive fertility impact of inactivity among men – after controlling for income. The latter effectively accounts for the loss of earner capabilities while not holding a job. The positive impact of joblessness among men on fertility appears counterintuitive, but has also been reported in previous studies (Kravdal, 2002; Lindo, 2010; Schmitt, 2012b; Inanc, 2015). The mechanism could be associated with a loss of male bargaining power, where being the breadwinner might enable women to promote their fertility intentions. The increased social acceptance of paternal childcare may also be contributing to this development: when fathers are able to take on a significant share of care responsibilities, family formation be- 35 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany comes a more feasible option, particularly in a welfare state context that makes the reconciliation of work and family difficult. A study on Sweden found that couples in which the man is unemployed tended to display a more equal division of labor (Ström, 2002). This study aimed to extend the literature on the employment-fertility nexus by honing in on the role of individual risk attitudes. In particular, the specific contribution of this study is to consider individual risk attitudes as possible moderators in the relation between employment uncertainties and fertility decisions. While there is a growing body of literature that links fertility behavior to personality traits, the investigation of risk attitudes, particularly in work contexts has been a blind spot. The findings suggest that risk aversion primarily speeds up the transition to parenthood. However, risk proneness also cushions the delaying effect of being long-term-trapped in uncertain employment contexts, as was shown in the case of marginal work. Whether the findings for are generalizable beyond Germany remains a question for future research. Concluding from this study, the German context poses a highly specific amalgam of a traditional gender ideology and an encompassing welfare state orientation. Recent policy changes in Germany include a shift in family policies that encourages female employment, while simultaneously dismantling welfare support with a more restrictive employment protection legislation and benefit system. In sum, this aggravates the threat emanating from employment uncertainties. This specific background makes it difficult to generalize the findings beyond a national context. A fruitful endeavor for future research might be a cross-national comparison of the impact of risk attitudes across different welfareand employment protection regimes. Do risk prone individuals – ceteris Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 36 paribus - tend to make the same fertility choices in both protected and unprotected employment regimes? Do individuals get accustomed to looming uncertainties in de-regulated labor markets? Honing in on these questions from a comparative perspective offers promising paths to promote the understanding on the relation between employment uncertainties and fertility behavior in what has become an intensely investigated field. 37 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany References Adsera, A. (2011). Where are the Babies? Labor Market Conditions and Fertility in Europe. European Journal of Population, 27, 1-32. Ahn, N., & Mira, P. (2002). A Note on the Changing Relationship between Fertility and Female Employment Rates in Developed Countries. Journal of Population Economics, 15, 667682. Andersson, G. (2000). The Impact of Labor-Force Participation on Childbearing Behavior: ProCyclical Fertility in Sweden during the 1980s and the 1990s. European Journal of Population, 16, 293-333. Andersson, G., & Scott, K. (2008). Childbearing dynamics of couples in a universalistic welfare state: The role of labor-market status, country of origin, and gender. Demographic Research, 17, 897-938. Baizán, P., Aassve, A., & Billari, F. C. (2003). Cohabitation, marriage, and first birth: The interrelationship of family formation events in Spain. European Journal of Population/Revue européenne de Démographie, 19, 147-169. Barbieri, P. (2009). Flexible Employment and Inequality in Europe1. European Sociological Review, 25, 621-628. Berninger, I. (2013). Women’s income and childbearing in context. Acta Sociologica, 56, 97115. Borghans, L., Duckworth, A. L., Heckman, J. J., & ter Weel, B. (2008). The Economics and Psychology of Personality Traits. Journal of Human Resources, 43, 972 - 1059. Box-Steffensmeier, J., & Jones, B. S. (2004). Event History Modelling. A Guide for Social Scientists. Cambridge: Cambridge University Press. Breen, R. (1997). Risk, Recommodification and Stratification. Sociology, 31, 473-489. Buhr, P., & Huinink, J. (2014). Fertility analysis from a life course perspective. Advances in Life Course Research, 21, 1-9. Butz, W. P., & Ward, M. P. (1979). The Emergence of Countercyclical U.S. Fertility. The American Economic Review, 69, 318-328. Cazzola, A., Pasquini, L., & Angeli, A. (2016). The relationship between unemployment and fertility in Italy: A time-series analysis. Demographic Research, 34, 1. Ciganda, D. (2015). Unstable work histories and fertility in France: An adaptation of sequence complexity measures to employment trajectories. Demographic Research, 32, 843. Cobb-Clark, D. A., & Schurer, S. (2012). The stability of big-five personality traits. Economics Letters, 115, 11-15. Comolli, C. L. (2017). The fertility response to the Great Recession in Europe and the United States: Structural economic conditions and perceived economic uncertainty. Demographic Research, 36, 1549-1600. Cygan-Rehm, K. (2016). Parental leave benefit and differential fertility responses: Evidence from a German reform. Journal of Population Economics, 29, 73-103. de Lange, M., Wolbers, M. H. J., Gesthuizen, M., & Ultee, W. C. (2014). The Impact of Macroand Micro-Economic Uncertainty on Family Formation in The Netherlands. European Journal of Population, 30, 161-185. Del Bono, E., Weber, A., & Winter-Ebmer, R. (2012). Clash of Career and Family: Fertility Decisions after Job Displacement. Journal of the European Economic Association, Forthcoming. Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 38 Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., & Wagner, G. G. (2005). Individual Risk Attitudes: New Evidence From a Large, Representative, Experimentally-Validated Survey. IZA Discussion Paper Series. Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., & Wagner, G. G. (2011). Individual Risk Attitudes: Measurement, Determinants, and Behavioral Consequences. Journal of the European Economic Association, 9, 522-550. Donkers, B. B. M., Arthur van Soest. (1999). Estimating Risk Attitudes Using Lotteries: a Large Sample Approach. Tilburg Center for Economic Research Discussion Paper. Donnellan, M. B., Conger, R. D., & Burzette, R. G. (2007). Personality Development From Late Adolescence to Young Adulthood: Differential Stability, Normative Maturity, and Evidence for the Maturity-Stability Hypothesis. Journal of Personality, 75, 237-264. Dupray, A., & Pailhé, A. (2017). Does employment uncertainty particularly impact fertility of children of North African immigrants in France? A gender perspective. Journal of Ethnic and Migration Studies, 1-24. Eckel, C. C., & Grossman, P. J. (2008). Men, Women and Risk Aversion: Experimental Evidence In C. Plott & V. Smith (Eds.), Handbook of Experimental Economic Results (pp. 10631078). Amsterdam: Elsevier Ekelund, J., Johansson, E., Järvelin, M.-R., & Lichtermann, D. (2005). Self-employment and Risk Aversion. Evidence from Psychological Test Data. Labour Economics, 12, 649-659. Fairlie, R. W., & Holleran, W. (2012). Entrepreneurship training, risk aversion and other personality traits: Evidence from a random experiment. Journal of Economic Psychology, 33, 366-378. Friedman, D., Hechter, M., & Kanazawa, S. (1994). A Theory of the Value of Children. Demography, 31, 375-401. Gash, V., & McGinnity, F. (2007). Fixed-term contracts—the new European inequality? Comparing men and women in West Germany and France. Socio-Economic Review, 5, 467-496. Gerster, M., & Lappegård, T. (2010). Mother's employment and fertility in Norway. Statistics Norway, Research Department. Discussion Paper. Giesecke, J. (2009). Socio-economic Risks of Atypical Employment Relationships: Evidence from the German Labour Market. European Sociological Review, 25, 629-646. Gnoth, C., Godehardt, D., Godehardt, E., Frank-Herrmann, P., & Freundl, G. (2003). Time to pregnancy: results of the German prospective study and impact on the management of infertility. Human Reproduction, 18, 1959-1966. Gottschall, K., & Schröder, T. (2013). " Familienlohn": zur Entwicklung einer wirkmächtigen Normierung geschlechtsspezifischer Arbeitsteilung. WSI-Mitteilungen, 66, 161-170. Green, F. (2008). Temporary Work and Insecurity in Britain: A Problem Solved? Social Indicators Research, 88, 147-160. Grotti, R., & Scherer, S. (2014). Accumulation of Employment Instability Among Partners— Evidence from Six EU Countries. European Sociological Review, 30, 627-639. Gutiérrez-Domènech, M. (2008). The impact of the labour market on the timing of marriage and births in Spain. Journal of Population Economics, 21, 83-110. Hagestad, G. O., & Call, V. R. A. (2007). Pathways to Childlessness. Al Life Course Perspective. Journal of Family Issues, 28, 1338. Hank, K. (2001). Changes in Swedish women’s individual activity status and the subsequent risk of giving birth in the 1980s and 1990s: An extension of studies by Gunnar Andersson and Britta Hoem. Demographic Research, 4, 125-132. 39 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany Hanoch, Y., Johnson, J. G., & Wilke, A. (2006). Domain Specificity in Experimental Measures and Participant Recruitment. Psychological Science, 17, 300-304. Hiilamo, H. (2017). Fertility Response to Economic Recessions in Finland 1991-2015. Finnish Yearbook of Population Research, 52, 15-28. Hobcraft, J., & Kiernan, K. E. (1995). Becoming a Parent in Europe. Prepared Paper: European Population Conference, September 4-8, 1995, Milano, Italy. Hoem, B. (2000). Entry into motherhood in Sweden: the influence of economic factors on the rise and fall in fertility, 1986-1997. Demographic Research, 2. Hondroyiannis, G. (2010). Fertility Determinants and Economic Uncertainty: An Assessment Using European Panel Data. Journal of Family and Economic Issues, 31, 33-50. Huinink, J., & Kohli, M. (2014). A life-course approach to fertility. Demographic Research, 30, 1293. Huttunen, K., & Kellokumpu, J. (2016). The effect of job displacement on couples’ fertility decisions. Journal of Labor Economics, 34, 403-442. ILO. (2016). Non-standard employment around the world: understanding challenges, shaping prospects. In (pp. xxv, 374 p.). Geneva: ILO. Inanc, H. (2015). Unemployment and the timing of parenthood: Implications of partnership status and partner's employment. Demographic Research, 32, 219. Kalleberg, A. L. (2009). Precarious Work, Insecure Workers: Employment Relations in Transition. American Sociological Review, 74, 1-22. Kravdal, O. (2002). The impact of individual and aggregate unemployment on fertility in Norway. Demographic Research, 6. Kreyenfeld, M. (2004). Fertility decisions in the FRG and GDR: An analysis with data from the German Fertility and Family Survey. Demographic Research, 3, 275-318. Kreyenfeld, M. (2010). Uncertainties in Female Employment Careers and the Postponement of Parenthood in Germany. European Sociological Review, 26, 351-366. Kreyenfeld, M., & Andersson, G. (2014). Socioeconomic differences in the unemployment and fertility nexus: Evidence from Denmark and Germany. Advances in Life Course Research, 21, 59-73. Kreyenfeld, M., Andersson, G., & Pailhé, A. (2012). Economic uncertainty and family dynamics in Europe: Introduction. Demographic Research, 27, 835-852. Kurz, K., Steinhage, N., & Golsch, K. (2005). Case Study Germany: Uncertainty and the Transition to Adulthood. In H.-P. Blossfeld, E. Klijzing, M. Mills & K. Kurz (Eds.), Globalization and Youth in Society (pp. 51-81). London: Routledge. Leibenstein, H. (1981). Economic Decision Theory and Human Fertility Behavior: A Speculative Essay. Population and Development Review, 7, 381-400. Liefbroer, A. C., & Corijn, M. (1999). Who, What, Where and When? Specifying the Impact of Educational Attainment and Labour Force Participation on Family Formation. European Journal of Population., 15, 45-75. Lindenberg, S. (1985). An Assessment of the New Political Economy. Its Potential for the Social Sciences and for Sociological Theory in Particular. Sociological Theory, 3, 99-114. Lindenberg, S. (1991). Social Approval, Fertilty and Female Labour Market Behaviour. In J. J. Siegers, Jenny de Jong-Gierveld & E. Van Imhoff (Eds.), Female Labour Market Behaviour and Fertility (pp. 32-61). Berlin, Heidelberg, New York, London, Paris, Tokyo, Hong Kong, Barcelona, Budapest: Springer. Lindo, J. M. (2010). Are children really inferior goods? Evidence from displacement-driven income shocks. Journal of Human Resources, 45, 301-327. Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 40 Lundström, K. E., & Andersson, G. (2012). Labor-market status, migrant status and first childbearing in Sweden. Demographic Research, 27, 719-742. Lutz, K. (2014). Compensating dissatisfaction in the job by turning to the family? The impact of current occupation on timing of first births in Germany. Advances in Life Course Research, 21, 43-58. Matysiak, A., & Vignoli, D. (2013). Diverse Effects of Women’s Employment on Fertility: Insights From Italy and Poland. European Journal of Population / Revue européenne de Démographie, 29, 273-302. Mayer, K. U. (2009). New Directions in Life Course Research. Annual Review of Sociology, 35, 413-433. McCrae, R. R., & Costa Jr, P. T. (1997). Personality trait structure as a human universal. American Psychologist, 52, 509. Mencarini, L., & Tanturri, M. L. (2006). High fertility or childlessness: Micro-level determinants of reproductive behaviour in Italy. Population, 61, 463-491. Meron, M., & Widmer, I. (2002). Unemployment Leads Women to Postpone the Birth of Their First Child. Population (English Edition, 2002-), 57, 301-330. Mynarska, M., Matysiak, A., Rybińska, A., Tocchioni, V., & Vignoli, D. (2015). Diverse paths into childlessness over the life course. Advances in Life Course Research, 25, 35-48. Neels, K., Theunynck, Z., & Wood, J. (2013). Economic recession and first births in Europe: recession-induced postponement and recuperation of fertility in 14 European countries between 1970 and 2005. International Journal of Public Health, 58, 43-55. Nicholson, N., Soane, E., Fenton-O'Creevy, M., & Willman, P. (2005). Personality and domainspecific risk taking. Journal of Risk Research, 8, 157-176. O'Rand, A. M. (2009). Cumulative processes in the life course. In The craft of life course research. (pp. 121-140). New York, NY, US: Guilford Press. Özcan, B., Mayer, K. U., & Luedicke, J. (2010). The impact of unemployment on the transition to parenthood. Demographic Research, 23, 807-846. Pailhé, A., & Solaz, A. (2012a). The influence of employment uncertainty on childbearing in France: A tempo or quantum effect? Demographic Research, 26, I. Pailhé, A., & Solaz, A. (2012b). The influence of employment uncertainty on childbearing in France: A tempo or quantum effect? Demographic Research, 26, 1-40. Reynaud, A., & Couture, S. (2010). Stability of Risk Preference Measures: Results From a Field Experiment on French Farmers. Rindfuss, R. R., Guzzo, K. B., & Morgan, P. S. (2003). The Changing Institutional Context of Low Fertility. Population Research and Policy Review, 411-438. Sahm, C. R. (2007). Stability of Risk Preference. Finance and Economics - Board of Governors of the Federal Reserve System Discussion Paper Series. Sahm, C. R. (2012). How much does risk tolerance change? The quarterly journal of finance, 2, 1250020. Santow, G., & Bracher. (2001). Deferment of the First Birth and Fluctuating Fertility in Sweden. European Journal of Population / Revue européenne de Démographie, 17, 343-363. Scherer, S. (2009). The Social Consequences of Insecure Jobs. Social Indicators Research, 93, 527-547. Schmidt, L. (2008). Risk Preferences and the Timing of Marriage and Childbearing. Demography, 45, 439-460. Schmitt, C. (2012a). A Cross-National Perspective on Unemployment and First Births. European Journal of Population, Forthcoming / Online First. 47 Economic Uncertainty, Precarious Employment, and Risk Attitudes in Germany Figures 3–6: Relative hazards – selected interaction plots Note: Estimation method for all hazard plots: Cox proportional hazard. Source: SOEP v33, 1990-2015, cohorts 1965-1989; own estimates. n=9 928 Figure 3: Relative hazards of first parenthood decisions. Risk attitudes by gender. n=4 732 Figure 4: Relative hazards of first parenthood decisions. Interaction between type of employment and risk attitudes Note: Atypical work summarizes employment in part-time and in marginal work. Omitted controls: Couple joint (log)net income, not working (as part of activity control set). Economic Uncertainty, Precarious Employment and Risk Attitudes in Germany 48 n=4.906 Figure 5: Relative hazards of first parenthood decisions. Interaction between risk aversion and duration in marginal employment Note: The risk-aversion dummy distinguishes between the more risk-averse (11-point scale values below 4) and neutral to more risk-prone individuals. n=4.906 Figure 6: Relative hazards of first parenthood decisions – women. Contour plots of risk tolerance and duration in marginal employment Legend: A risk tolerance of 5 points (on a 0 to 10 scale, with 0 being highly risk-averse), and a duration of 6 months in marginal employment exert the same relative hazard on the propensity to opt for parenthood as a risk tolerance of 2 points and a duration of 30 months in marginal employment. .6 .8 1 1.2 1.4 Predicted Hazard Ratio 0 12 24 36 Duration in Marginal Employment (in months) Risk Averse Not Risk Averse Predictive Hazard Rate of Marginal Employment - Women