Analyzing Women s Employment and Fertility Rates in Europe: differences and similarities in Northern and Southern Europe

Transcription

1 Analyzing Women s Employment and Fertility Rates in Europe: differences and similarities in Northern and Southern Europe Daniela Del Boca* daniela.delboca@unito.it Silvia Pasqua** silvia.pasqua@unito.it Chiara Pronzato* chiara.pronzato@unito.it 5/29/2003 Keywords: Labor Market Decisions, Fertility. J.E.L Subject Codes: J2, C3, D1 (*) Università di Torino - Dipartimento di Economia Via Po 53, Torino, ITALY - Phone: Fax: and CHILD (Centre for Household, Income, Labour and Demographic Economics) (*) Università di Torino - Dipartimento di Scienze Economiche e Finanziarie, Corso Unione Sovietica 218bis, Torino, ITALY - Phone: Fax: and CHILD (Centre for Household, Income, Labour and Demographic Economics) 1

2 1. Introduction The decline in fertility and the growth in women s labor market participation in most OECD countries has been one of the most important economic and demographic events of the last decades. Several important changes characterize the cross-country and temporal pattern of participation and fertility. The correlation between female participation and fertility, that had been negative and significant for several years, has changed sign since the 1980s and has became positive and weaker, while important differences emerge among countries. In order to explain the observed negative relationship between fertility and income, economic models have introduced women s allocation of time decisions and emphasized the opportunity costs of women s time. The increase in women s education, relative earnings and employment rates, for example, has increased the opportunity costs of childbearing (Becker 1964, Willis 1973). More recently, to explain the changes in sign and significance of the relationship between participation and fertility, economists have stressed the role of labor market characteristics, the availability and costs of childcare as well as the patterns of marriages and divorce (Ahn and Mira 2002, Billari et al 2002). In Southern European countries a positive correlation has been observed between low participation and low fertility, while in Northern European countries a positive association emerges between high participation and high fertility. Because of these different characteristics, the research focusing on the relationship between women s participation and fertility has developed especially in Southern European countries, where conciliating work and fertility is still very difficult. On the contrary, in Northern European countries, where fertility and participation are more compatible, economic research has focused on the effects of children on wages and careers. In this paper we want to analyze women s decisions concerning work status and fertility, considering both individual and household characteristics, and also the effects of labor market characteristics and availability of childcare services. Moreover, we will analyze women s labor market behavior in the period surrounding the event of childbirth. In section 2 we discuss the analytical framework that has been used to study labor supply and fertility. In section 3 we present important facts that are likely to have affected the recent trends. In section 4 we review several empirical results concerning the relationship between women s work status and fertility in various countries. In section 5 we present a theoretical model to analyze women s fertility and participation. In Section 6 we describe the dataset used (ECHP, European Community Household Panel), while in section 7 and 8 we present our empirical results. 2

3 2. The basic theory of the relationship between fertility and women s labor participation For a long time economists have neglected the analysis of fertility and its determinants, basically because of the difficulties of incorporating it rigorously into the traditional theory of the consumer choice. Fertility was essentially considered a non-economic phenomenon, it was seen as a socio-demographic one, thus the analysis of fertility was outside the scope of the economic theory. Only when economic theory started to focus on human capital and allocation of time, were fertility variables (as other demographic, sociologic and psychological variables, such as marriage, divorces, education, etc.) introduced among the other traditional economic variables, such as income, consumption and saving (Browning 1992). The modern literature in the area is based upon the work of Becker (1964), who asserts that fertility could be analyzed within an economic framework emphasizing the connection between income and fertility. Assuming that children are normal goods, the demand for children will increase with the household income. Since husbands are not usually involved in childcare activities, an increase in the husband s wage has the same positive effect on the demand for children as an increase in the household income. On the contrary, since childrearing is a mother s time intensive activity, increasing wives wages could have a negative effect on the demand for children (Willis 1973). In other studies the quality and quantity of children have been considered. Income increase may reduce fertility if the income elasticity for the quality of children is sufficiently greater than the income elasticity for the quantity (Becker and Lewis 1973, Willis 1973). More recently, fertility and labor market participation decisions have been recognized as the joint result of the maximization of household expected lifetime utility under budget and time constraints in a dynamic context. The desired participation status and the desired number of children depend on the whole sequence of prices and wages and on a variety of characteristics reflecting preferences. The two types of decisions are therefore simultaneous in the sense that they are the solution to a common constrained maximization problem (Cigno 1991, Rosenzweig and Wolpin 1980, Hotz and Miller1988, Francesconi 2002, Del Boca 2002). The increasing use of panel data allows researchers to take into account the dynamics involved in the relationship between births and women s work status. Moreover, it makes it possible to take into account important omitted factors, such as fecundity, tastes, and other individual and marriage-specific traits, which are important factors in explaining the decision to have children and are unobservable to the researcher. Finally, the most recent directions in 3

4 household economics have underlined the limits of traditional models based on the unitary approach, where household income is pooled, and have analyzed fertility and women s labor supply decisions as the results of the interactions between family members. Indeed, it seems that fertility, as well as women s labor supply, are the object of a bargaining process between the spouses (Del Boca 1997, Schultz 1998). Understanding the relationship between fertility and labor supply is critical to a number of policy debates. Recent studies have analyzed the relationship between labor market participation and fertility by considering not only the effect of prices and incomes and the household characteristics but also several aspects of the labor market and the social service system (part time, unemployment, childcare system, and parental leave) (Del Boca 2002, Bratti 2001). In our empirical study we try to consider both women s and men s personal characteristics and variables allowing us to analyze the effects of social and labor market policies. 3. The facts The most recent data indicate that two important patterns have characterized the relationship between fertility and participation in most OECD countries: while for many decades the relationship had been negative, in the late 1980s the correlation became positive and weaker 1. Even if a decline in fertility and an increase in the participation pattern have been observed in almost all industrialized countries, important differences emerge in the levels, suggesting that different countries are in different stages of development and are constrained by specific social and economic factors (Table 1). While in Southern European countries both fertility and participation are low, in Northern European countries both fertility and participation are high. Figure 1 illustrates these differences. Some interpretations have suggested that the weakening link is mainly due to the growth in the unemployment rates and to a greater availability of childcare (Ahn and Mira 2002, Ermisch 1989, Hotz et al. 1997). Important differences characterize the labor markets and welfare systems of Northern and Southern countries. In Northern countries part-time employment is very widespread and represents most of the opportunities offered to women. On the contrary, it is extremely rare in Southern European countries, and it is an important factor accounting for the low employment rates of married women, particularly those with children (Table 3). 1 While the correlation was 0.5 during the 1970s and the beginning of the 1980s, by the late 1980s the correlation became

5 Unlike the North of Europe, the Southern European labor markets are still characterized by strict rules regarding the hiring and firing of workers and the employment arrangements available. These labor market regulations have been largely responsible for the high unemployment rates of women and youth. Table 2 shows unemployment rate in different countries. The hiring system and the high entry wage as well as very strict firing rules severely restrict employment opportunities for labor market entrants: because entry-level positions are so hard to find, many children live with their parents until they find their first stable employment (Table 4). Thus the labor market indirectly imposes high fertility costs on families even when the mother does not work, and therefore it discourages fertility both directly and indirectly. Moreover, by encouraging women not to leave the labor market when they have young children, but to maintain continuous attachment to their job, it has a negative effect on fertility (Table 5). Another important aspect concerns the hours worked by women: compared to other European countries, in the Southern European countries such as Italy, Spain and Greece female employees work longer hours (Table 6). The slow change in the division of labor between men and women in Southern European families has been an additional constraint with reference to working and having children. While in Northern European countries social protection in the welfare regime is generous in services, in Southern European countries it is relatively generous in transfers, but it continues to rely on the role of the family as producer and distributor of services. The most important example of the lack of social services concerns the availability of childcare for children under 3 years of age (Table 7). In Southern Europe childcare services are typically inadequate and characterized by extreme rigidity in the number of weekly hours available. This makes the service compatible with part-time work, but not with full-time activities. Having school-age children does not necessarily increase the attractiveness of full-time employment, since school days often end in mid-afternoon, thus making childcare necessary for late afternoon and early evening. Where part-time is limited, married women are forced to choose between no work or full-time work, neither of which is necessarily their preferred option. Married women who choose to work tend to have full-time work commitments, which is not compatible with having a high number of children. The differences described above contribute to explain the different shapes of participation rates in the life cycle of women in different countries (Figure 2). 5

6 4. Recent empirical studies The increase in the proportion of parents at work has stimulated an intense debate concerning the importance of factors affecting work and fertility. Personal characteristics and family support The changes in women s education affect wages as well as wages profiles, with important effects on participation and fertility decisions and the timing of the events (Gustafsson 1994, 1995). Bratti (2001) found that education raises the job attachment of women. In particular, highly educated women work also in the period surrounding a birth event and therefore education induces fertility postponement. These results imply that policies aiming at increasing women s education would have a positive effect on participation but uncertain effects on fertility, given evidence of a U-shaped pattern of fertility with education (interpreted in terms of the prevalence of income over substitution effects due to education as well as to the circumstance that highly educated women have more access to the expensive private childcare system). When only mother s time and market goods are required for childrearing, an increase in working women s wages induces both income and substitution effects on fertility (and on labor supply). An increase in wages leads to an increase in labor supply at low wages, where the substitution effect dominates the income effect. At higher wages it may be the opposite. Ahn and Mira (2002) argue that the increasing importance of income effects when female wages increase could yield to a change in the cross-country correlation between fertility and women s participation over time. Ermisch (1989) provides empirical evidence of this for Britain. As already mentioned, the most recent directions in household economics have analyzed the limits of traditional unitary models and the importance of using collective models to analyze household decisions. The implications of this change in the theoretical approach to household decision-making processes are particularly relevant for economic policies: the effects of public policies (taxes and transfers) on intra-household resource distribution are not neutral, in fact they influence the behavior and the demand for leisure of husbands and wives differently, depending on who is the beneficiary (Del Boca 1997, Schultz 1998). 6

7 Labor market characteristics Empirical studies using cross-country data discuss the possibility that the emergence of high and persistent rates of unemployment in Europe might have contributed to the acceleration of the fertility decline and might have been responsible for the reversal in the sign of the correlation between fertility and participation (Ahn and Mira 2002). Expectations of high and persistent unemployment have different effects on fertility: on the one hand, women will participate in the labor market to protect household income against negative shocks to partners wage and employment; on the other hand, they will not leave employment during childbearing years to protect their won labor market prospects. Del Bono (2001) directly tests the hypothesis that expectations of future labor market outcomes affect current fertility decisions. Her results show that unemployment is one of the variables that most significantly affect the expectations of future wages and job opportunities, and it may therefore be responsible for the decline in fertility. Other studies have more closely considered the relationship between fertility and labor market characteristics. Bettio and Villa (1998) focus their attention on the way how women form their expectations of future labor market conditions and project these expectations on to their fertility decisions. High unemployment rates may affect fertility in a significant manner. Not only does the experience of unemployment reduce current income, it also affects the level of income that families consider necessary for the wellbeing of their children. Another important characteristic of the labor market concerns the availability of part-time jobs. Comparative studies have found a high correlation between the proportion of part-time jobs and the participation rates of women, in particular married women with children (Meulders et al. 1994). The low proportion of part-time workers does not seem to be coherent with self-reported preferences. A large number of women who are unemployed or out of the labor force report that they would actually prefer to work part-time. Surveys at different points of time and in different areas have reported similar results (European Economy 1995). Bardasi and Gornick (2000) show that being a mother (compared with being childless) decreases the probability of selecting full-time work and increases the probability of both nonworking and working part-time. Also Tanda (2001) analyses the impact of part-time using ECHP data and finds that the availability of part-time jobs increases the probability for a women to be employed. Analyzing the relationship between motherhood and careers prospects Guetierrez- Domènech (2002) shows that the career transition after a first birth may be of different types: 7

8 women who used to work may become either unemployed or inactive after it or may experience downward occupational mobility. That is, even if a woman remains employed she may end up in an occupation that is below the one held before the birth in terms of quality, payment and responsibilities. There are several potential reasons why this happens. On the one hand, mothers may be willing to supply labor that involves fewer responsibilities as long as they take care of their children. On the other hand, employers may be reluctant to hire mothers for high profile positions since their family role may absorb their energy and interfere with their productivity. As we mentioned above, youth unemployment has implied an additional burden on Southern European families and mothers. Southern European youth looking for a job or for a more stable source of income cohabit with their families much beyond mature age. The role of the family in support of the children often extends far beyond the completion of schooling by the children. Because of the limited access to credit and housing markets on the part of individuals without stable employment, the Southern family traditionally provides income support to its children during their usually lengthy search for a stable, "protected job. This responsibility is likely to have significant effects on women s participation and fertility (Giannelli and Monfardini, 2002, Martinez-Granado Castillo, 1998, Billari et al 2002). Family friendly policies Economic theory suggests that a parental leave policy would result in more women remaining in the labor force, since they do not have to leave employment to take time off from work, but it may lead to more women on leave at any points in time and thus to fewer employed. Table 8 summarizes maternity leave policies in European countries. When examining the changes in labor supply of newborns mothers in countries where maternity leave statutes were passed between 1980 and 1990, Klerman and Leibowitz (1994) did not find a significant higher employment rate and greater leave taking, but they did find evidence for a lower rate of newborns mothers actually being in their jobs in countries that had passed maternity leave legislation. A second expectation is that parental leave may reduce efficiency if it encourages women - who would otherwise be employed - to stay out of work longer. On the other hand, more women might enter employment if they knew they have access to leave. Ruhm and Teague (1997) examined the association between leave policies and indicators of macroeconomic conditions, and found that paid leave is associated with an increased domestic GDP, increased 8

9 employment and reduced unemployment. Unpaid leave is associated with increased labor force participation and employment, but increases unemployment presumably because unpaid leave is not sufficient to discourage those mothers who want longer leaves from leaving their jobs. Tanda (2001) compares the effects of both compulsory and optional maternity leave regulations in the European countries on female employment rates. A long compulsory maternity leave period has a negative impact on the probability of women to work, possibly because employers perceive this as a negative effect of hiring a female worker, but also for the discouraging effect on women s decision to participate. On the contrary, the length of optional maternity leave has a positive effect on women s employment rate. The availability of childcare services significantly affects women's preferences for nonmarket time versus time spent in paid work. Improvements in childcare options as well as variations in their costs have been associated with significant increases in the labor supply of mothers in most countries (Gustafsson and Stafford 1992, Gustafsson 1994, 1995). Ermisch (1989) findings have suggested that the increased availability of market childcare is one possible explanation for the change in fertility. According to his results, the changing effect on fertility of increases in women s wages (from negative to positive) would be facilitated by the availability of market childcare. The decisions to work and to have a child are both positively influenced by the availability of childcare. Childcare availability seems to be more important than its costs. Del Boca (1993, 2002) has analyzed the effect of childcare on the participation of married women in Italy, and the specific characteristics of the supply of public and private childcare systems. Using different datasets the studies reach similar conclusions, that is, household labor supply depends on childcare rationing rather than its costs. The estimate of the relationship between childcare costs and labor supply shows that a reduction in childcare costs increases the probability of mothers to work parttime, but has no effect on the probability of working full time (Del Boca 1993). In Italy, childcare costs seem to affect labor supply only in areas where places are available. Where childcare availability is limited and its cost is high, a high number of working mothers have to rely on family support systems (mainly to grand-parents help) or to babysitters, since also the private childcare system has still a limited offer (Del Boca Locatelli and Vuri 2003). 9

10 5. A model of fertility and participation decisions In analyzing both fertility and participation we try to take into account not only personal and family characteristics, but also characteristics related to the institutional environment (labor market, service system). Let be w it a variable indicator which takes the values = 1 if the woman i works at time t, = 0 else. Especially for the analysis of the fertility decision one important problem derives from the omission of factors (such as fecundity, taste and marriage characteristics). To take into account the unobservable individual specific factors we use a fixed effects model with panel data. The fixed effects logit estimator for the decision of working is given by: p(w it =1 η i X it ) = [exp(η i +βx it )] / [1+ exp(η i +βx it )] i=1, N; t =1, T; where η represents the unobservable characteristics (which, we suppose, do not vary over time) and X are the observable characteristics, which may or may not vary over time, T=4. Using a fixed effects model with panel data we can take into account and isolate the effect of these factors. If we assume that η is independently and identically distributed as a normal random variable, we use a random effects estimator; otherwise we use a fixed effects estimator. The random effects estimator is more efficient, but may be inconsistent if η is not independently distributed with respect to the observed covariates. The problem with the fixed effects estimator is that it allows to estimate only the effects of those variables that change over time (Chamberlain 1980). To decide which estimator should be considered more reliable we test the restriction that households do not systematically differ in terms of any measured variables, using standard Hausman tests. First we consider both decisions simultaneously: the fertility decision (f) and the participation decision (w): p(w it =1, f it =1 η W i, η F i, X W it, X F it ) = p(w it =1 η W i X W it ) * p(f it =1 η F i X F it ). Given the independence between the decisions f and w conditional on the X and the η, the estimator for each decision is independent of the estimator for the other decision. For details on the econometric model characteristics see Del Boca

11 6. The data Our empirical analysis utilizes a four-year panel data from the European Community Household Panel (ECHP), a standardized multi-purpose longitudinal survey coordinated and supported by Eurostat, which allows to study and compare the Member States of the European Union. The survey involves annual interviewing of a representative panel of households and individuals in each country, covering a wide range of topics on living conditions such as income, employment, poverty and social exclusion, housing, health, migration and other social indicators. The unit of analysis are the families and, within the households, all individuals older than 16. Nevertheless, it is possible to get information about all family members, including children under sixteen. We select women in the age range 21-45, married (or cohabitant) from several European countries: Italy, Spain, Belgium, France, United Kingdom and Austria for the years The dependent variables are whether the wife is working at the time of the interview and whether or not she has had a child in the last two years. The sample size for the participation equation is cases in the fixed effects model and cases in the random effects model. For the fertility equation the sample size is in the fixed-effects model and in the random-effects. The independent variables are related to: Personal characteristics: wife s age, education, non-labor income; Family characteristics: husband s total income, age of the eldest child and number of children already born in the household; Labor market characteristics: regional unemployment rate, ratio of part time jobs to the employment by region.; Childcare services: regional percentage of children (0-3 years old) who attend a pre-primary school. (for the year 1999 only). Only a few of the independent variables are not time-invariant. The information concerning income is made comparable (using PPP) using specific coefficients provided by Eurostat in the ECHP data. Unemployment rates and percentages of part-time jobs, for different European regions, are taken from REGIO, a Eurostat dataset providing regional data. The percentage of children (0-3 years old) who attend a pre-primary school, region by region, is only available for the year 1999 (from different sources). We use these data in a cross-section analysis for the year The table in the Appendix describes the variables used in the econometric analysis. 11

12 7. Participation and fertility decisions: empirical results Tables 10 and 11 report the fixed and random effect estimates of employment and fertility probability in the first and second column respectively. We can compare the random and fixed effects of common parameters when the coefficient is associated with a time-varying variables. First, we compare fixed and random effect results in the employment equation (Table 10). The wife s age has a positive impact on participation but is statistically significant only in the random effect model. Finally, wife s schooling has the expected positive sign on participation (this coefficient cannot be estimated using the fixed effects estimator). Both wife s non-labor income and husband s income have negative effects on employment probability. Also the composition of the family has an important effect. The number of children already present in the family has a negative effect on women s participation in the labor market, while the age of the eldest child has a positive but not significant effect. The environmental variables have a significant impact: the regional rate of unemployment has a negative effect on the probability to work, while the availability of part-time jobs has a positive effect on the probability to work. Living in a Southern European country reduces the probability of wife s working. Childcare availability - which is included only in the cross-section equation - has a positive and significant effect. The cross-section estimates are quite similar in magnitude and sign to the panel results. The Hausman test statistics reported at the bottom of Table 10 indicate rejection of the null hypothesis of the equality of all unobservables across households. In the fertility equation (Table 11) the woman s age has a positive effect on the probability of having children, which is coherent with the postponement phenomenon. The positive schooling effect (again this coefficient cannot be estimated using the fixed effects estimator) can be interpreted as a permanent income effect, given that the father s education is not included in the analysis in presence of assortative mating (Del Boca 2002). The number of children already present in the household has a negative effect on the probability of having an additional child, while the variable age of the eldest child has a positive and significant impact. Wife s non-labor income has a positive effect, while husband s total income has a positive but not significant effect. Regional unemployment rate has a negative but not significant effect, while the effect of part-time is positive and significant. Living in the South of Europe has no significant impact. Finally, childcare availability has a positive but non significant effect on fertility. The 12

13 Hausman test statistics reported at the bottom of Table 11 indicate rejection of the null hypothesis of the equality of all unobservables across households. This implies that the fixed effects model has to be preferred. These results indicate that several institutional characteristics affect the joint decision of working and having children. In particular, in areas where unemployment rates are higher women find it more difficult to work and have children, while in areas where part-time jobs are available the opposite is true. The availability of childcare increases the probability of both events, although it has a significant effect only on the probability of participating. 8. Women s work status around childbirth: empirical results Now we turn to analyze women s work status changes after a childbirth, and specifically after the first childbirth. We consider three possibilities of work status: working full-time, working part-time and not working at all. We examine the transition probability from one work status to another. We restrict our sample to the sub-sample of those women who had their first childbirth at time t and we consider two types of transitions: from working at time t-1 to not working at time t+1, and from working full-time at t-1 to working part-time at t+1. The transition probability from a work status (W) to a non-work status (NW) after the first childbirth is given by: Prob (transition from W to NW) = (NW t+1 W t-1 ) /(W t-1 ) where NW t+1 is the number of women who are not working one year after their first childbirth, and W t-1 is the number of women who are working one year before their first childbirth. The probability to change from a full-time job (FT) to a part-time job (PT) is given by: Prob (transition from FT to PT) = (PT t+1 FT t-1 ) /(FT t-1 ) We select all women who have their first child during the period and then we can observe their work status one year before and one year after the first childbirth. Table 12 shows that in all countries (but Portugal) women work less hours after their first childbirth. Figure 3 shows that only in The Netherlands, Belgium and Ireland the probability to change from a full-time activity to a part-time activity after the first childbirth is higher than the probability to leave the labor market. 13

14 In Southern European countries a smaller number of women change their status after childbirth. As we have discussed above, the proportion of women working in these countries is much lower than in the rest of Europe, and they are prevalently employed in full time permanent jobs. Using monthly information on work status, we can also calculate the proportion of women working around their first childbirth, month by month, from twelve months before to thirty months after childbirth (see Gustafsson et al. 2002). In this case, since dataset information is not so detailed, we cannot distinguish between part-timers and full-timers. Figure 4 shows that in Spain and Italy the percentage of working women does not seem to change as a consequence of childbirth, while we have a U-shaped curve around the month of the childbirth, while in France the decline is six months afterwards, as a consequence of the Allocation Parentale d Education (APE) that gives a strong incentive for the mothers to leave the labor market, especially when they have relatively low wages or precarious jobs (Périvier and O Dorchai 2001). 14

15 Bibliography Ahn, N., and Mira P., (2002), A note of the relationship between fertility and female employment rates in developed countries Journal of the Population Economics, 14(Spring) Albrecht, J. A., Edin, P.A., Sundström, M., and Vroman, S., (1999), Career Interruptions and Subsequent Earnings: A Reexamination Using Swedish Data," (M.), Journal of Human Resources Bardasi E. and Gornick J. C., (2000), Women and Part-Time Employment: Worker Choices and Wage Penalties in Five Industrialised Countries, ISER Working Paper Becker, G., (1964), Human Capital, N.Y. Columbia U. Press Becker, G., (1981), A Treatise of the family, Harvard University Press Becker, G., Lewis, H.G., (1973), On the Interaction between the Quantity and Quality of Children", Journal of Political Economy, 81 Bettio, F., Villa, P., (1998), A Mediterranean Perspective on the Break-down of the Relationship between Participation and Fertility Cambridge Journal of Economics, 22, Billari F. The Emergence of lowest Low Fertility in Europe during the 1990.Population and Development Review. 28 (4) : Bratti, M., (2001), Labour Force Participation and Marital Fertility of Italian Women: the Role of Education, paper presented at the AIEL Conference, October 2001 Browning, M., (1992), Children and Household Economic Behavior, Journal of Economic Literature, Vol.XXX, Carliner, G., Robinson, C., and Tomes, N., (1980), Female Labour Supply and Fertility in Canada, Canadian Journal of Economics, 13(1), Cigno, A., (1991), Economics of the Family, Oxford University Press, Oxford Colombino U. and Di Tommaso M., (1996) ``Is the Preference for Children so Low or is the Price of Time so High?'', Labour, 10(3), Del Boca, D., (1997) Intrahousehold Distribution of Resources and Labor Market Participation Decisions, in Economics of the Family and Family Policies (Persson, I., Jonung C. eds ) Routledge Press Del Boca, D., (2002), The Effect of Child care and Part Time on Participation and Fertility of Italian Women, Journal of Population Economics, 14 Del Boca D, (2002b) Labor market Behavior and Child care Opportunities in Monitoring Italy ISAE Rome, 2002 Del Boca D., Locatelli M. and Pasqua S., (2000), Employment Decisions of Married Women: Evidence and Explanation, Labour, 14, Del Boca D., Locatelli M., Pasqua S. and Pronzato C., (2003), Data base on motherhood, WP MOCHO, Del Bono, E., (2001), Estimating Fertility, Working Paper 2001 University of Oxford, Pembroke College Delemarre N. (2001), kinderopvang in gemeenten. De monitor ove 1999, report 15

16 Di Tommaso M., (1999) A Trivariate Model of Participation, Fertility and Wages: the Italian Case, Journal of Cambridge Economics, 23, Ermisch, J., (1989), Purchased Child Care, Optimal Family Size and Mother s Employment: theory and Econometric Analysis, Journal of Population Economics, 2, Ermisch, J., Francesconi, M., (2001), The effects of parents employment on children s lives, Family Policy Studies Centre European Economy, (1995), Performance of the EU Labor Market: Results of an ad hoc Labor Market Survey, European Commission B-1049 Brussels Eurostat (1999), Demographic Statistics Francesconi M. (2002) A Joint Dynamic Model of Fertility and Work of Married Women Journal of Labor Economics 20, Giannelli G., Monfardini C., 2002, The probability of cohabitation of adult children, University of Florence Gornick, J.C., Meyers, M.K., Ross, K. E., (1997), Supporting the Employment of Mothers: Policy Variation across Fourteen Welfare states, Journal of European Social Policy, 7, Gornick J. C., Meyers M. K. and Ross K. E., (1998), Public Policies and Employment of Mothers: A Cross-National Study, Social Science Quarterly, 79, Gustafsson, S., (1994), Childcare and Types of Welfare States, in D. Sainsbury Gendering Welfare States pp Thousand Oaks, Ca: Sage Gustafsson, S., (1995), Public Policies and Women's Labor Force Participation (P. Schultz ed.) Investments in Women Human Capital, Yale University Press Gustafsson, S., Stafford, F., (1992), Daycare Subsidies and Labor Supply in Sweden, Journal of Human Resources, 27(1), Gustafsson S.and Wetzels C., (2000), Optimal age at first birth :Germany Great Britain the Neherlands and Sweden in Gustafsson S. and Meulders D. E Gender and the Labour Market Mac Millan London Gutièrrez-Domènech, M., (2002), Job Penalty after Motherhood: A Spanish Case in a European Context, Family Friendly Policies Conference IZA Conference, May, Bonn Hotz, V.J., Miller, R.A., (1988), An Empirical Analysis of Life Cycle Fertility and Female Labor Supply, Econometrica, 56(1), Hotz, V.J., J.A. Klerman, and R. J. Willis, (1997), The Economics of Fertility in Developed Countries: A survey, In M.R. Rosenzweig and O. Stark, editors. Handbook of Population and Family Economics, Vol. 1A, Chapter7, , North Holland, Amsterdam ISFOL (2001), Rapporto trimestrale. Klerman J. A., Leibowitz A, (1994), Labor Supply Effects of State Maternity Leave Legislation in F. D. Blau Ronald G. Gender and Family Issues in the Worklace New York Russel Sage Foundation Kunze, A., (2002), Timing of Birth and Wages paper presented Family Friendly Policies Conference, IZA Bonn Lesthaeghe, R., and Willems, P., (1999), Is Low fertility a Temporary Phenomenon in the European Union?, Population and Development Review, 25(2),

17 Martinez-Granado M. Castillo,(1998), The participation decisions of Youth, WP98-2, Department of Economics Carlo III, Madrid Meulders, D., Plasman, O., Plasman R., (1994), Atypical Employment in the EC, Aldershot. Moreno Luis, 2001, Superwomen and Mediterranean Welfare mimeo University of Oviedo OECD (1999), Employment Outlook OECD (2000), Employment Outlook OECD (2001), Employment Outlook Périvier H. and O Dorchai S. (2002), Women s Employment and Public Policies, State of the Art, Report for the European Union of the Project MOCHO (The Rationale of Motherhood Choices: Influence of Employment Conditions and of Public Policies) Rosenzweig, M., Wolpin, K.I., (1980), Life-Cycle Labor Supply and Fertility: Casual Inferences From Household Models, Journal of Political Economy, 88(2), Ruhm C. J., (2002), Parental Employment and Children welfare, paper presented Family Friendly Policies Conference, IZA Bonn Ruhm C. J., Teague J.L., (1997), Parental Leave Policies in Europe and North America in F. D. Blau Ronald G. Gender and Family Issues in the Worklace New York Russel Sage Foundation Schultz T. P.,(1998), Eroding the economic foundations of marriage and fertility in The United States in Structural Change and Economic Dynamics, Special issue on The Economics of the family (Del Boca D. editor), Noth Holland, Vol. 9 Tanda P., (2001), ISAE Report on Politiche Sociali: Selettivita e Universalismo United Nations, 1995, Human Development Report Wetzels, C, (2002), Does Motherhood Really make Women Less Productive? The case of the Netherlands, Bilbao ESPE Conference Willis, R.J., (1973), A New approach to the Economic Theory of Fertility Behavior, Journal of Political Economy, 81(2),

18 Table 1. Fertility and female participation rates in European countries Fertility rates Participation rates (%) (*) 2000 Nordic countries Denmark Finland Sweden Central Europe Austria Belgium France Germany Ireland Luxembourg Netherlands U.K Southern Europe Greece Italy Portugal Spain Source: Eurostat (2001), Statistics in Focus (*) OECD (1997), Employment Outlook Table 2. Unemployment rates in European countries 1983(*) 2000 Nordic countries Denmark Finland Sweden Central Europe Austria Belgium France Germany 7.7 (**) 7.9 Ireland Luxembourg Netherlands U.K Southern Europe Greece Italy Portugal Spain

19 Source: Eurostat (2001), Statistics in Focus (*) OECD (1997), Employment Outlook (**) West Germany Table 3. Proportion of women working part time and in the service sector (2000) Par- time workers Employed in the service sector Nordic countries Denmark Finland Sweden Central Europe Austria Belgium France Germany Ireland Luxembourg Netherlands U.K Southern Europe Greece Italy Portugal Spain Source: Eurostat (2001), Statistics in Focus Table 4. Children living with their parents in 1987 and 1995 (%) Denmark Sweden France U.K Greece Italy Spain Source: Eurostat (1997) 19

20 Table 5. Youth unemployment rate, total fertility rate and female employment rate (2000) Youth unemployment rate Total fertility rate Female employment rate Nordic countries Denmark Finland Sweden Central Europe Austria Belgium France Germany Ireland Luxembourg Netherlands U.K Southern Europe Greece Italy Portugal Spain Source: Eurostat (2001), Statistics in Focus Table 6. Average hours worked by women per week (2000) Weekly hours Nordic countries Denmark 31.9 Finland 35.5 Sweden 33.9 Central Europe Austria 33.9 Belgium 31.3 France 33.4 Germany 31.3 Ireland 32.2 Luxembourg 33.6 Netherlands 24.5 U.K Southern Europe Greece 38.0 Italy 34.1 Portugal 36.1 Spain 35.9 Source: our elaborations on Eurostat (2001), Statistics in Focus 20

21 Table 7. Child care for children under 3 and 3-6 (%) Aged under 3 Aged 3-6 Social protection expenditure on family/ children (in % of GDP) Nordic countries Denmark Sweden Central Europe Austria Belgium France Germany Netherlands 18(*) U.K Southern Europe Greece Italy Spain Source: OECD (2001), Employment Outlook, and our elaborations on Eurostat (2001), Statistics in Focus (*) Delemarre (2001) Table 8. Maternity leave policies in Europe Base maternity leave Duration (weeks) Benefits (% of average wages) Optional maternity leave Duration (weeks) Benefits (% of average wages) Nordic countries Denmark Sweden Central Europe Belgium France Germany Netherlands U.K Southern Europe Greece Italy Portugal Spain Source: Tanda (2001) 21

22 Table 9. Participation equation Fixed effects Random effects Cross section 98 age,021,046**,033** secondary school - -,796** -,782** less than secondary school - -1,624** -1,649** wife s non-labor income -,020** -,043** -,031** husband s total income -,003 -,009** -,011** number of children -,335** -,288** -,238** age of the oldest child,019,014**,022** unemployment -,083** -,047** -,042** part time,038** -,037** -,035** South - -,860** -,888** childcare -,042** costant - 1,510** 1,461** Hausman test 21,3 **=significant at 95%; *=significant at 90% Table 10. Fertility equation Fixed effects Random effects Cross section 98 age,202** -,076** -,075** secondary school - -,150** -,146 less than secondary school - -,162** -,113 wife s non-labor income,052**,094,067** husband s total income,000,001** -,000 number of children -5,781** -,913** -1,072** age of the oldest child,163** -,200** -,214** unemployment -,012,022**,029** part time,270**,019,029** South - -,065,004 childcare -,012 costant - 3,834** 3,801** Hausman test 1686,6 **=significant at 95%; *=significant at 90% 22

23 Table 11. Average hours worked by women per week around their first childbirth Before After Nordic countries Denmark Central Europe Austria Belgium France Germany Ireland Netherlands U.K Southern Europe Greece Italy Portugal Spain Source: our elaborations on ECHP Table 12. Participation after the first childbirth Fixed effects Random effects Cross section 98 Age -,090,053**,052** secondary school - -,745** -,894** less than secondary school - -1,027** -1,256** attitude - 2,806** 2,487** wife s non-labor income -,024** -,025 -,118** husband s total income,010,000 -,010* months elapsed,041**,017**,011 second child -2,018** -1,045** -,817** unemployment -,239** -,012,009 part time -,187**,038*,054* South -,015 -,149 childcare -,096** costant - -2,660** -3,160** Hausman test 48,0 **=significant at 95%; *=significant at 90% Figure 1 23

24 Women s participation and fertility (2000) 80 participation rate Austria Germany Sweden Portugal Denmark Finland UK The Netherlands France Belgium Ireland Luxembourg 40 Spain Greece Italy ,25 1,5 1,75 2 TFR Source: our elaborations on ECHP Figure Figure Source: Eurostat (2001), Statistics in Focus Women s participation rates by age Italy Great Britain Sweden EU e oltre 24

25 Transitions in Europe around first childbirth percentage of women D DK NL B F UK IRL I EL ES P A countries from full to part time jobs from in to out of the labor market Source: our elaborations on ECHP Figure 4 Participation around first childbirth 1 0,9 percentaga of women working 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0, months B F UK I E Source: our elaborations on ECHP 25