Does Female Employment Always Undermine Marriage? Working Wives and Family Stability in Different Contexts of Italian Society
by Lorenzo Todesco
University of Turin
Sociological Research Online, 17 (3) 13
Received: 15 Aug 2011 Accepted: 14 Mar 2012 Published: 31 Aug 2012
Previous research has shown that the association between female employment and risk of marital disruption is still far from clear-cut, partly because certain theoretical and empirical evidence indicates that it may vary according to different conditions. The purpose of this study is to reassess the association between female employment and marital stability in Italy, by viewing it as contingent on historical period, institutional and cultural context and wives' gender ideology. The relative risk of marital disruption is estimated using discrete time event-history models. The empirical findings clearly show that wives' employment in this country seems to be disruptive for marriages, and its effect remains constant across the different conditions tested in the analysis.
Keywords: Family - Italy - Marital Dissolution - Divorce - Female Employment - Working Wives
Introduction1.1 Conventional knowledge suggests that spousal sharing of the role of economic provider weakens the institution of marriage. Actually, the association between female employment and risk of marital disruption is both complex and controversial. In recent decades, the transformation from a manufacturing to a service-based economy and the shift in marital role expectations have resulted in a continuing increase in female labour force participation in industrialized societies (Oppenheimer 1994). Since this increase is paralleled by a rise in marital disruption, the association between wives' employment and marital stability has become a common topic in sociological, economic and demographic research analysing the determinants of divorce. In general, the results of individual-level studies are far from being univocal (see e.g., Ruggles 1997; Schoen et al. 2006, coming to opposite conclusions). Moreover, some theoretical arguments supported at the empirical level maintain that this association may vary according to different conditions (Greenstein 1995; South 2001; Cooke & Gash 2010). In Italy, however, no research has yet been conducted to shed light on this topic. The study presented herein attempts to fill this gap, reassessing the association between wives' employment and marital stability by viewing it as contingent on three different factors, namely the historical period, the institutional and cultural context and wives' gender ideology. The choice of this country makes the investigation particularly interesting because of certain peculiarities distinguishing Italy from other Western countries as regards the female employment pattern and divorce behaviour. In these respects, Italy provides an innovative and interesting setting because the subject-matter of the study has never been analysed in a country marked by a combination of a very low female employment rate and high marital stability. There are other reasons that make this study interesting from a methodological standpoint. The study uses a dataset whose longitudinal retrospective sections and large sample size make it particularly suitable for analysing how the association between female employment and marital stability can vary according to different conditions. Moreover, the richness of the dataset in terms of variables also makes it possible to control for a number of theoretically important determinants of both marital disruption and wives' employment, thus avoiding spurious associations.
1.2 The study is organized as follows. The first section addresses the subject-matter of the study by detailing the theoretical framework and providing a description of the empirical studies conducted in various Western countries. The second section focuses on the Italian case. At the end of this section, three hypotheses are proposed for testing in the rest of the study. The third section outlines the dataset and the research methodology. The fourth section presents the findings of the empirical analysis, and the fifth section discusses their implications. The final section is devoted to some brief concluding remarks.
Theoretical and Empirical Evidence2.1 Several competing theoretical frameworks provide predictions about the potential impact of female employment on marital stability. The hypothesis of a disruptive impact can be derived from several perspectives, most of which are included in the so-called specialization and trading model. With extensive roots in both micro-economic (Becker et al. 1977; Becker 1981) and sociological theories (Parsons 1949, 1955), these perspectives support the contention that a traditional division of labour within the marital couple decreases the risk of disruption, but they have done so for different reasons. In Becker's view, marriage is a highly efficient strategy for both men and women when one partner specializes in market human capital, i.e., those skills that increase earning capability in paid work, and the other specializes in non market human capital, i.e., those skills related to managing domestic tasks as well as child rearing. As a consequence, if both spouses devote time to paid work, specialization is reduced, the gains to marriage decline and, in turn, the desirability of staying married (see also Weiss 1997). Because of the greater male attachment to the labour force, and the generally higher wages, the most stable marriages are those in which the husband exchanges economic support for wives' household tasks, and vice versa.
2.2 In the 1950s, Parsons maintained that sex-role segregation is a functional necessity for marital stability and, more in general, for the survival of society as a whole. According to this scholar, a sexual division of labour is required in order to assure full effectiveness for the family as a social system. Because women give birth to children, it seems natural that they should care for them, as well as for the house, the husband and, more generally, for family relations, while men are mainly involved in their working careers (Parsons 1955). Parsons (1949) also maintains that in urban and industrial societies the marital relationship is structurally unsupported and, as such, is unstable and requires certain protective mechanisms in order to last (in this respect see also Goode 1961, 1962). The main mechanism reducing disruptive competition between the spouses is sex-role segregation, where the dominant male role is occupational and the dominant female role is that of housewife and mother. It follows that wives' employment on the one hand lessens role differentiation, specialization and interdependence, and on the other increases the competition between wives and husbands, which in turn undermines marital stability. Remaining in the field of sociology, more recently Cherlin (1992) argued that women with an independent income are less disposed to devote time and energy to working out marital problems.
2.3 Although the specialization and trading model is often considered elegant and persuasive, it has not received full support from the empirical analysis of the impact of indicators of female labour force supply on marital stability (e.g., employment status, marital work experience, number of hours spent in paid employment). Focusing on the United States, some studies report a positive association between female employment and risk of marital disruption in accordance with expectations (Booth et al. 1984; Spitze & South 1985; South & Spitze 1986; Ruggles 1997; Hiedemann et al. 1998; Brines & Joyner 1999), others find only modest or qualified support (Greenstein 1990; Tzeng & Mare 1995), or no association at all (Bumpass et al. 1991; Presser 2000; Sayer & Bianchi 2000; Rogers 2004). Still other studies raise serious doubts about this model, pointing out that the entry of women in the workforce has weakened marriage not because of the direct effect of female employment, but indirectly as a result of changes in the pool of available partners (South & Lloyd 1995), by influencing wives' earnings (Ono 1998) or only in the case of unhappy marriages (Schoen et al. 2002). A recent study by Teachmann (2000) underlines the complexity of the association between female employment and marital stability: on the one hand, wives' cumulative labour force participation has a positive effect on marriage, on the other an increase in wives' work hours is associated with a rise in the risk of disruption.
2.4 In other Western countries, the specialization and trading model has received more straightforward support than in the United States. However, it must be emphasized that studies are meagre in number for each country and usually not devoted entirely to the analysis of the association between female employment and marital stability, and caution is consequently required in drawing conclusions. There is some evidence that female employment has a disruptive effect in Australia (Bracher et al. 1993), Sweden (Hoem & Hoem 1992; Trussell et al. 1992), Norway (Tjøtta & Vaage 2003), the United Kingdom (Chan & Halpin 2002), Spain (Coppola & Di Cesare 2008) and Hungary (Bukodi & Róbert 2003), a modest effect in Finland (Jalovaara 2003), and an inconsistent effect in Flanders (Corijn 1999, cit. in Kalmijn et al. 2004: 78). In the Netherlands, some studies find that wives' work is positively related to divorce probabilities (De Graaf & Kalmijn 1999, cit. in Poortman & Kalmijn 2002: 177; Fokkema & Liefbroer 2004; Kalmijn & Poortman 2006), but others do not (Manting 1993, Pit & Rouwendaal 1994, cit. in Poortman & Kalmijn 2002: 177).
2.5 Besides the mixed findings given by the empirical evidence, the specialization and trading model and its arguments have been strongly criticized by Oppenheimer (1994, 1997), for both theoretical and methodological reasons. This scholar puts forward a different theoretical framework, often called the flexibility model, to shed light on the association between female employment and marital stability. The assumptions of the specialization and trading model rely on male family wages and on lifelong permanent employment, two features of the labour market that are increasingly unlikely in post-industrial service-based economies (Guillemard 2005). In Oppenheimer's view, a strong sex-role specialization in marriage is a high-risk and inflexible family strategy, because the temporary or permanent loss of one specialist can mean that functions vital to the well-being of the complementary specialist and offspring are not being performed. Specialization may be a feasible strategy only in a large extended family household where no particular individual is indispensable because of the redundancy in personnel characterizing such a system (in this sense see e.g. Czap 1982). Moreover, sex-role specialization is also an inefficient way to deal with the varying needs of nuclear families over their developmental cycle. In these families, the ratio of consumers to producers and consequently the family's level of living varies substantially over time (Lee 1983), and specialization involves a potentially serious inflexibility in dealing with the changes in a family's internal composition (for a comparative view on these see Kuijsten 1996). By contrast, women's employment enhances family economic flexibility, which should lend greater financial security and in turn less strain and more marital stability. Oppenheimer also argued that the correlation between female employment and marital stability is largely a function of the starting point chosen, almost invariably sometime in the 1950s or early 1960s. If the time series is pushed farther back, it is clear that the divorce rate had been rising for many decades before female employment started its rapid growth.
2.6 The flexibility model has received some support by a few recent empirical analysis (Bramlett & Mosher 2002; Schoen et al. 2006) carried out in the United States, according to which wives' labour force participation does not inherently weaken, but rather may strengthen marriage.
2.7 The specialization and trading model and the flexibility model postulate opposite directions in the association between women's employment and risk of marital disruption, but implicit in both is the assumption that this association tends to be invariant across historical periods, institutional and cultural contexts and individual-level characteristics. However, there are good reasons for suggesting otherwise.
2.8 As regards historical period, although no current theoretical framework of marital dissolution specifically predicts that the effect of female employment has changed over time, South (2001) maintains that some competing arguments support both the hypothesis of a reduction of its magnitude and that of an increase. On the one hand, the destabilising effect of women's work on marriage may be expected to decrease because of the cultural and economic changes surrounding such work. When relatively few wives work in the paid labour force, those who do violate traditional gender role expectations, and this violation is likely to be disruptive for marriages (see e.g. Rogers & Amato 2000). The recent acceptance of married women's employment may mitigate this disruptive effect. In addition, the rapid increase in women's job opportunities means that even nonworking wives contemplating marital dissolution can expect to find gainful employment should their marriages dissolve (in this respect see Raeymaeckers et al. 2008; Van Damme et al. 2009). When female employment rates were low, nonworking wives had fewer prospects of getting a job in the event of marital disruption, and this difficulty may have rendered them more likely to remain in unhappy marriages. On the other hand, several circumstances may have strengthened the negative association between female employment and marital stability. The expansion of social and family policies may have made it increasingly easy for working women to cope following the end of the marriage, by providing for family leaves, childcare services and flexible non-standard work schedules. The relative lack of such facilities in a earlier period may have made even employed women reluctant to dissolve their marriages, limiting the difference in the risk of marital dissolution between working and nonworking wives (Pagnini & Rindfuss 1993; Presser 2000). Moreover, following the liberalization of gender role attitudes, women in the paid workforce have less need to compensate for their atypical economic arrangement by emphasizing more traditional behaviours in other aspects of the marital relationship, for instance by avoiding using their economic power to dissolve unhappy marriages (see the so-called gender display (Brines 1994) and deviance-neutralization (Greenstein 2000) patterns). Finally, because of the historical decline in occupational sex segregation (for an overview see e.g. Jarman et al. 1999), employed married women increasingly work in close proximity with men who are potentially more attractive mates than their current husbands (in this respect see also McKinnish 2007).
2.9 In his empirical analysis performed in the United States, South (2001) maintains that the effect of wife's hours worked on marital stability has become increasingly disruptive over the observation period. By the periods 1969-1976 and 1977-1984 no effect is found, while by the period 1985-1992 employed wives seem to have less stable marriages than unemployed ones; the predicted annual probability of marital disruption among the former exceeds that for the latter by about 40%. However, these findings are not supported by other studies carried out in Australia (Bracher et al. 1993), in the Netherlands (Poortman & Kalmijn 2002) and in Germany (Beck & Hartmann 1999, cit. in Poortman & Kalmijn 2002: 178; Wagner 1997, cit. in Kalmijn et al. 2004: 86), showing that the destabilising effect of female employment has decreased over time.
2.10 Shifting attention to the effect of the institutional and cultural context on the association between female employment and marital stability, recent cross-national longitudinal studies (e.g. Cooke 2006) have provided valuable insights. Cooke and Gash (2010) maintain that what proves disruptive to marital stability is not an individual wife's employment per se, but when her engagement in paid work differs from the norm of the society. In this sense, a prominent role is played by social and family policies whereby female labour force supply can be supported or discouraged (e.g. Gornick & Meyers 2003). When more women are in the workforce, female employment becomes more normative and the risk of marital disruption associated with it should be mitigated. Because the normative level varies between and within countries according to the institutional and cultural contexts, the effect of female employment on marital stability should also vary accordingly. Other scholars have applied a normative argument to the increase in women's educational attainment (Härkönen & Dronkers 2006), in cohabitation (Liefbroer & Dourleijn 2006; Wagner & Weiß 2006) and in parental divorce (Lyngstad 2006) on attenuating the greater risk of marital disruption that these individual factors historically predicted.
2.11 A recent comparative study (Liefbroer & Dourleijn 2006) provides clear empirical evidence to the argument stating a variation in the association between female employment and marital stability according to the institutional and cultural context. This study reveals some important cross-countries differences: being employed increases the risk of union dissolution in Finland, West Germany, Austria, Italy, Poland and Lithuania, it lowers the risk in France and Latvia, and no effect is found in Sweden, Norway, Flanders, Spain, East Germany, Hungary, the Czech Republic and Slovenia. Unfortunately, the implications of these findings are not discussed by the study's authors. Another comparative study including Great Britain, West Germany and the United States performed by Cooke and Gash (2010) shows that couples where the wives worked full-time do not face the risk of disruption more than those where the wives were out of the labour force in any of these countries. Because of the high female employment rate of all three countries, this lack of difference in terms of marital stability between working and nonworking wives supports the normative argument. Moreover, West German marriages were most stable in the case of couples where the wives worked part-time, with a 43% lower risk of marital dissolution than that of couples where the wives were out of the labour force. Among British and U.S. couples, wives' part-time employment did not alter the risk of disruption. These findings are also in line with the normative prediction, since part-time employment in West Germany is supported by social and family policies, and a large proportion of part-time jobs are professional public-sector positions that do not marginalize workers. As a consequence, most West-German part-time workers do not incur the same wage penalty as those from Great Britain and the United States (in this sense see also Kalleberg et al. 2000; Warren 2000), and a substantial percentage of employed women works part-time, reflecting an increase in recent years.
2.12 Focusing on individual-level characteristics, Greenstein (1995) devotes some attention to how and why women's gender ideology might moderate the effect of female employment on the stability of the marriage. His argument starts from the theoretical framework developed by Major (1987) and Thompson (1991) on the perceptions of fairness. It is a well-known fact that husbands often do not share the housework burden equally even if their wives are employed full-time in the labour market (for a recent review of the studies supporting this statement see Lachance-Grzela & Bouchard 2000). Assuming that female employment detracts from marital stability also because of the conflicts generated in fulfilling the traditional gender division of work and responsibilities (in this respect see e.g. Voydanoff & Donnelly 1999; Stohs 2010; Frisco & Williams 2003), women with traditional gender attitudes might show a weaker positive effect of their employment on the risk of disruption than nontraditional women. The former tend to accept the doctrine of the segregated marital roles, where the ideal marriage consists of an economic successful male breadwinner and a dependent nurturing wife; consequently, these women view the inequality in the division of household labour as consistent with their ideology, and therefore not necessarily unfair. On the contrary, nontraditional women find this inequality onerous and unfair, and their reactions may result in decreased marital quality and, in the long term, decreased marital stability.
2.13 This theoretical model is of interest since it drives scholars to devote attention to the possible interaction of gender ideology in the association between female employment and marital stability. However, it gives great importance to individual resources and gender roles, but it does not take into account the recent theorizations analysing with a more fluid approach the dimensions of gender and power within a couple relationship. Gender scholars are increasingly raising doubts on the notion that the basis of power is predominantly material (Sassler & Miller 2010). Instead, a crucial role in decision-making process is played by the social norms regarding appropriate behaviours for men and women and by the social institutions limiting possibilities for deviating from these behaviours (Ferree 1990; Martin 2004; Risman 2004). Komter (1989) suggests the concept of hidden power in order to explain the decision-making processes in which conflict does not take place because the parties follow the hegemonic norms of what is considered appropriate and expected. Other scholars (Carli 1999; Ridgeway 2001) maintain that the existing social structures legitimate and perpetuate the belief associating greater worthiness and competence with men than women. Within this stream of theories, a valuable insight comes from the scholars (e.g. Heimer & Staffen 1995; Martin 1998; McGuffey & Rich 1999; Miller & Sassler 2010) referring to the so-called "doing gender" approach put forward by West and Zimmermann (1987). This theoretical framework confutes the idea of gender as a social role characterized by a sort of script carefully taught and learned in which distinct and integrated attitudes and behaviours are stored. On the contrary, gender can be rather understood as "a routine accomplishment embedded in everyday interactions" (West & Zimmerman 1987: 125). Gender is produced on the everyday level through ongoing behavioural displays, which are imbued with symbolic significance for gender: men and women perform different tasks because such practices affirm and reproduce gender selves, thus reproducing a gender interaction order. Moreover, men and women differ in their degree of masculinity and femininity and need to be constantly reminded to be masculine and feminine; in other words, men and women have to "do" gender rather than "be" a gender (for an overview on gender roles and doing gender see Fox & McBride Murry 2000). Unfortunately, these theories on gender and power cannot be tested in a quantitative analysis like that put forward here, which is confined to assess if the association between female employment and marital stability varies according to a proxy variable capturing the gender ideology of the respondents (for more details, see the methodological section).
2.14 At the empirical level, the research devoted to the effect of gender ideology on the association between female employment and marital stability gives mixed findings. Greenstein (1995) maintains that in the United States there is no effect of wives' hours per week in paid employment on marital dissolution for women holding a traditional gender ideology, while a disruptive effect is found for nontraditional women. Nontraditional women who were not employed during marriage had about a 44% lower risk of marital dissolution compared with those employed 20-35 hours per week, and those employed 35-40 hours had about two-thirds the risk than did women employed more than 40 hours. However, another study performed in the Netherlands by Kalmijn, De Graaf and Poortman (2004) produced opposite findings: female employment substantially increases the odds of divorce for traditional women, while for nontraditional ones no differences between working and non-working women are found in terms of marital stability.
The Italian Case3.1 The goal of this study is to analyse the association between female employment and marital stability in Italy and how it varies across historical periods, institutional and cultural contexts and wives' gender ideology. This country provides a challenging and original case of study for shedding light on this association because of a combination of distinctive features connected with these issues. On the one hand, marriages are still more stable than in most European countries. At the institutional level, evidence in this sense comes from the legislation regulating divorce proceedings, which is inspired by very strict principles: to obtain a divorce, a couple must wait at least three years after the decree of legal separation. Thus, legal separation is the first, essential step towards formally ending a marital relationship, and divorce is granted only some years later. For this reason, the appropriate indicator of the magnitude of marital instability in Italy is based on the number of legal separations, rather than on the number of divorces, even because many legally separated couples never go as far as divorce (more details on this issue are supplied in the methodological section). Focusing on divorce behaviour, when the Italian crude separation rate (the number of legal separations per 1,000 population) is compared with the crude divorce rate (the number of divorces per 1,000 population) of the other 27 European Union countries, only Ireland, Greece, and Slovenia show lower rates. On the other hand, focusing on the female employment pattern, it must be emphasised that in the typology of welfare states developed by Lewis and Ostner (1995) in the mid-Nineties, Italy was classified among the strong male breadwinner countries. Hence, it goes without saying that women were not a large part of the workforce: quite the opposite. Nowadays, things have not changed much in this respect. When the Italian female employment rate is compared with that of the other 27 European Union countries, only Malta shows a lower rate. Italy also shows a very low percentage of couple families where both adults are in paid employment if compared with certain other European countries (OECD 2011). In sum, in Italy a very low female employment rate goes hand in hand with high marital stability, and the subject-matter of the study has never been analysed in a country with such characteristics. Moreover, Italy has a strongly traditional vision of the family in terms of formation procedures, family and gender roles, and obligations among family members, as shown by studies which analyse values and attitudes towards the family (e.g. Künzler 2002). As regards the aim of the study, it is a matter of interest to shed light on the change in the association between female employment and marital stability across women's gender ideology in such a conservative country.
3.2 The association between female employment and marital stability, like all issues that relate, broadly speaking, to legal separation and divorce, have received little scholarly attention in Italy. The few studies carried out so far analyse the effect of wives' paid work together with several other determinants of marital dissolution. No study has been fully devoted to an in-depth analysis of the association between female employment and marital stability. After the pioneering study by De Rose (1992) based on data from the beginning of the Eighties, this association has only recently been tackled by studies controlling for both marital duration and other confounding variables (Arosio 2004; Coppola & Di Cesare 2008; Todesco 2009; Vignoli & Ferro 2009; Salvini & Vignoli 2011). All these studies show that working wives seem to have less stable marriages than nonworking wives, lending some support to the specialization and trading model.
3.3 A shortcoming in the scarce Italian literature is the implicit assumption that the stability of all marriages are affected in the same manner and at approximately the same level of magnitude by the wife's employment. However, those studies presented in the previous section analysing interactions of wives' employment with other variables have tended to find that the effects of the former on marital stability do vary according to the latter. Thus, the present study increases the current body of knowledge, testing the hypothesis according to which in Italy as elsewhere, the effects of wives' employment on marital stability may be moderated by, or interact with, a host of other factors.
3.4 It is important to stress that the effect of the institutional and cultural context on the association between female employment and marital stability and the normative argument can be empirically tested in the analysis presented here even if this is a single country study. This is due to the fact that a characteristic of the Italian labour market is the large within-country variation in women's participation. In the Northern part of the country, the female employment rate exceeds 56%, while in the Southern part it is below 31%. These areas show different normative levels of female employment, as well as large differences in cultural values and meanings surrounding wives' paid work, and differences in terms of family gender roles. Moreover, there is a substantial gap in social services and policies aiming at supporting working mothers at regional and local level. Hence, the different areas of the country provide different institutional and cultural contexts for observing the changes in the association between female employment and marital stability.
3.5 Like most of the studies carried out so far, the present study is confined to testing the association between wives' employment and marital stability, assuming that the effect goes from wives' employment to marital stability and not the opposite. A well-founded alternative interpretation is that the risk of marital disruption increases wives' labour market participation, because they anticipate the end of the marriage (e.g. Thompson & Amato 1999). Testing this interpretation would require a different type of analysis from that presented here (e.g. Johnson & Skinner 1986; Rogers 1999). It follows that a part of the effect of wives' employment on marital stability might be endogenous. However, a recent study by Poortman (2005) tests the hypothesis of the anticipatory behaviour comparing the effect of wives' work on divorce between divorces varying in the extent to which they were expected. According to her line of reasoning, the effect of the anticipatory behaviour should be cancelled out for unexpected divorces, because women were not able to make labour market adjustments in anticipation of the end of the marriage. The findings show that the disruptive effect of full-time employment is relatively strong also when the divorce was fully unexpected. Thus, Poortman concludes that female employment increases the divorce risk over and above the anticipatory behaviour.
3.6 In the following sections, three hypotheses derived from the previous theoretical and empirical studies are tested:
H1: Female employment is expected to have different effects on the risk of marital disruption contingent upon historical period, in terms of direction and/or magnitude.
H2: The differences between working and nonworking wives in terms of marital stability should be smaller in the Northern than in the Southern part of the country, because of the institutional and cultural context that makes female employment more normative and promoting gender equality.
H3: The impact of female employment on the risk of marital dissolution differs according to wives' gender ideology: the effect is expected to be less disruptive for traditional than for nontraditional women.
Data and Methods4.1 The data for this study come from the Families and Social Subjects Survey performed by the Italian National Institute of Statistics (ISTAT), a large dataset that constitutes the Italian part of the Gender and Generation Survey. The survey was conducted in 2003 and consisted of personal interviews with a national sample of 49,541 respondents (19,227 households). Most sections of the survey were cross-sections, but some longitudinal retrospective sections covered the life course (mainly marital and fertility histories) and working career. This dataset is well suited to analysing this study's subject matter, both because it contains longitudinal retrospective information on marital and working careers, and because it provides a wealth of data on a number of theoretically important determinants of marital disruption. Moreover, its large sample size facilitates the estimation of statistically reliable parameters for the interaction effects of interest. The initial subsample used in the analysis comprises 12,652 female respondents, of whom 11,683 are at first marriage and 969 have obtained a legal separation or divorce. With the restrictions due to the deletion of cases with missing data for critical variables in the analysis, this study includes 11,780 female respondents.
4.2 The effect of female employment on marital stability and its changes are estimated using event-history analysis, in this specific instance some discrete time event-history models. The estimating method is an ordinary binomial logistic regression which estimates, for individual i, the odds of experiencing marital dissolution at time t as:
where Pit is the hazard of the event occurring at time t for an individual with time-constant covariates Xit and time-varying covariates Wit, Zit is a function of the duration for individual i at time t and β0, β1, β2, and β3, are unknown parameters. The use of robust standard errors solves the problem of correlated observations because there is more than one observation for each individual. The dependent variable is a time-varying variable assuming value 1 in the event of legal separation or zero if the marriage lasts. The observation period to the risk of marital dissolution is measured from the year of marriage to the year of legal separation. Unfortunately, many divorced respondents cannot be included in the analysis, because the year of their legal separation is not available. If a marriage remains intact, the observation period is measured from the year of marriage to the year of the interview. The observation period is broken down into a series of 12-month intervals in the analysis. Some limitations in the data prevent consideration of shorter intervals. As pointed out by some scholars (Barbagli 1990; Barbagli & Saraceno 1998), in Italy the analysis of marital instability needs to be based on legal separations rather than on divorces, because of the strict legislation regulating the procedure for marriage dissolution already mentioned. Since divorce is granted at least three years after the decree of legal separation, considering the former rather than the latter implies two severe deficiencies. Firstly, marital instability would be underestimated, since many legally separated couples never go as far as divorce for various reasons: because people prefer not to have anything more do with the former spouse and consequently avoid the divorce proceedings, but also because people want to avoid renegotiating the arrangements made at the moment of legal separation or risk losing their marital rights (above all, inheritance and survivor pension). Secondly, the event of divorce follows at least of three years the first formal step to dissolve a marriage, and consequently it is very far away in time from the effective break-up of the marital couple. This seriously biases an event history analysis concerning a determinant of marital dissolution like that presented here. It must be underlined that the event of legal separation does not coincide with the effective break-up of the marital couple either, but the variable concerning the year of de facto separation is seldom available in the Families and Social Subjects Survey.
4.3 The analysis presented here is in line with those conducted in previous research on the subject-matter of this study. The key hypotheses guiding this analysis are concerned with how the effect of female employment on marital stability varies by historical time, institutional and cultural context and wives' gender ideology. Therefore, the regression models include the appropriate interaction terms testing for these hypothesized interaction effects. Four separate models were estimated in this study. In the first model, the main independent variable consists of a categorical time-varying variable measuring the employment status, coded 1 - "Employed", 2 - "Not employed". The category "Employed" includes respondents working without interruptions during the year and those that interrupted employment at least once. No differences are found between the former and the latter in terms of marital stability (table not shown, available on demand). This is the only information available in the dataset about the respondents' working career that can be reconstructed retrospectively across the duration of the marriage. A deficiency of this study is that no information about the respondents' income is stored in the dataset. For this reason, it is not possible to disentangle the direct effect of female employment on marital stability from the indirect one mediated through wives' earnings (in this sense see Ono 1998). The second model replicates the first and includes the interaction terms between the variable measuring the employment status and a categorical time-varying variable measuring the calendar period. The third model replicates the first and includes the interaction terms between the variable measuring the employment status and a categorical time-constant variable regarding the area of residence of the respondent. As regards the effect of wives' gender ideology on the association between female employment and marital stability, unfortunately a time-varying variable measuring such ideology is not available in the dataset. In absence of this information, an acceptable proxy may be whether or not the couple cohabited before marriage. I am conscious of the fact that this is a serious defect in the analysis, but no better solutions have been found. However, it must be stressed that many scholars point out the strong link between nontraditional gender ideology and de-facto unions (e.g. Lesthaeghe & Surkyn 1988; Rindfuss & VandenHeuvel 1990; Axinn & Thornton 1992; Thomson & Colella 1992; Clarkberg et al. 1995; Lye & Waldron 1997; Kaufman 2000; Sassler & Goldscheider 2004), and this is particularly true in a country as conservative in terms of family customs as Italy, where living together out of wedlock has long been rare (in this sense see Nazio & Blossfeld 2003) and is slowly becoming more normative only among very recent cohorts. Moreover, another argument supporting the link between nontraditional gender ideology and cohabitation stems from several studies (e.g Cunningham & Antill 1994; South & Spitze 1994; Baxter 2005; Davis et al. 2007) showing a more egalitarian division of unpaid family labour in cohabiting couples than in marital ones, with male cohabiters devoting more time than husbands to household tasks, even if female cohabiters still contribute more than their partners. The cross-national study by Batalova and Cohen (2002) also shows that couples' premarital cohabitation experience appears to contribute to greater equality in the division of household tasks during the subsequent marriage. Thus, the fourth model presented here replicates the first and includes the interaction term between the variable measuring the employment status and a categorical time-constant variable regarding the experience of cohabitation before marriage. Of course, a woman can have a nontraditional gender ideology even without having experienced a cohabitation before marriage. Moreover, gender ideology may change during the life course, according to ideas and situations at which individuals are exposed (in this sense see e.g. Bolzendahl & Myers 2004; Davis 2007). For these reasons, the effect of gender ideology on the association between female employment and marital stability may not be properly estimated in this analysis.
4.4 In addition to the variables concerning the association between female employment and marital stability and its variations, several control variables were included in the models to take into account of some important effects influencing both working career and marital stability as well, and thus avoid spurious associations: birth cohort, education, rite of wedding (which may capture an effect of gender ideology which is not captured by the experience of cohabitation before marriage) and number and age of children. All of these are categorical time-constant variables, except for number and age of children, which is a categorical time-varying variable.
4.5 A limitation of this study has to do with the lack of some relevant control variables. In general, there is agreement at the theoretical level on the fact that the risk of marital disruption is determined jointly by both spouses' characteristics (on this see e.g. Lyngstad 2004). Thus, it would be of interest including in the analysis some control variables referring to the respondent's husband, such as education, gender ideology and income, which may play a role in the association between wife's employment and marital stability. In particular, it would be necessary to control for both spouses' gender ideology, which may be theoretically conceptualized as a gendered lens linked to the wider gender system through which female employment is perceived. Gender ideology defines expectations about appropriate performance of male and female marital roles, and there is evidence both at the theoretical and at the empirical level that the effect of female employment on marital dissolution is affected by both spouses' gender ideology (see the ideological consistency model advanced by Ross and Sawhill (1975) and the empirical study carried out by Sayer and Bianchi (2000)). Unfortunately, this study is forced to adopt an individual-level rather than a couple-level perspective, since in the dataset employed here no information about the former spouse is available for respondents with an experience of marital dissolution. In other words, the subject-matter of the study is tackled only from the wife's side. However, it must be underlined that the husband's gender ideology is, at least to some extent, taken into account in the analysis. This due to the fact that wife's gender ideology is captured through the experience of cohabitation before the marriage. Since this is a couple-level variable, it captures some of the husband's gender ideology as well. In the same line of reasoning, husband's gender ideology is also captured by the variable referring to the rite of wedding.
4.6 The overall composition of the final subsample used in the analysis is presented in the appendix (Table A1), which contains observations (marriage-years) and events (legal separations) for each of the variables used in the modelling procedure.
Results5.1 Table 1 presents the results for the association between female employment and marital stability and its changes across historical periods, institutional and cultural contexts and wives' gender ideology from the discrete time event-history models, as well as the results for the effects of the control variables. This study does not go into detail for the latter, as these effects are in agreement with the findings of other studies conducted in Italy cited above.
|Table 1. Effects of Calendar Period, Area of Residence and Wives' Gender Ideology on the Association between Female Employment and Marital Stability and Control Variables|
5.2 In line with these studies, model 1 clearly shows that a marital disruption-promoting effect of wives' employment is found in this country: the odds that nonworking women end their marriages is 42% lower than the corresponding odds for working women. Considering the effect of the historical period on the association between female employment and marital stability, model 2 reveals that before the second half of the Eighties, the odds that nonworking women face a marital disruption was 25% lower than the corresponding odds for working women, became 41% lower in the period 1985-1989, and was also lower in the periods 1990-1994, 1995-1999 and 2000-2003, by respectively 57%, 39% and 46%. However, all interaction terms are very far from statistical significance. Accordingly, it can be stated that the effect of female employment on marital disruption in Italy remains constant over time. Dwelling on temporal dynamics, an explorative analysis points out that this effect does not change over marital duration (table not shown, available on demand).
5.3 Focusing on the association between female employment and marital stability in the different institutional and cultural contexts of Italian society, model 3 shows that in the Northern part of the country the odds that nonworking women dissolve their marriages is 40% lower than the corresponding odds for working women. No large difference is found in Central Italy (35% lower), while in Southern Italy the difference in the odds between nonworking and working women increases up to 52% lower, but the interaction term fails to reach statistical significance. Consequently, the normative argument is not confirmed by these data, and the destabilising effect of female employment on marriage seems to go beyond the differences in terms of labour market structure, social and family policies as well as of values and attitudes.
5.4 Finally, as regards the effect of wives' gender ideology on the association between female employment and marital stability, model 4 reveals that the odds that nonworking traditional wives (at least according to their cohabiting behaviour) disrupt their marital unions is 42% lower than the corresponding odds for their working counterparts. Shifting attention to nontraditional wives, no appreciable difference is found (44% lower). The fact that the interaction term is null in magnitude, apart from its statistical significance, raises some doubts concerning the hypothesis that the lack of variation of the association between female employment and marital stability across wives' gender ideology is due to a possible defect in the estimation of the latter (see the methodological section).
5.5 In sum, the association between female employment and marital stability in Italy seems not to interact with the different factors tested here. This outcome is supported not only by the fact that the interaction terms are always very far from statistical significance and sometimes null in magnitude, but also by the fact that the inclusion of such terms does not lead to an appreciable improvement in the fit of the models.
Discussion6.1 Following the changes in female labour force participation and divorce behaviour recorded in recent decades in most Western countries, a considerable amount of research has been performed to test different indicators of female labour force supply as determinants of marital disruption. However, both theoretical and empirical research have given mixed findings. Some scholars point out that the association between female employment and marital stability is controversial partly because it may vary according to a host of conditions. This study follows this stream of research, focusing on Italy and analysing the variation of this association across historical periods, institutional and cultural contexts and wives' gender ideology.
6.2 The findings of this study are far from being equivocal. Firstly, it seems clear that female employment has a strong disruptive effect on marriage in Italy. This is in line with the other studies carried out in this country as well as in a large part of Europe, even if it contrasts with the less clear-cut results of those performed in the United States. These findings are not unexpected for various reasons. Italy is well-known for being a country with limited family policies, specifically those devoted to labour and family reconciliation, e.g., parental leaves, childcare services, flexible time schedules, part-time working. It thus appears that the costs of this reconciliation fall primarily on the family, or rather on women's shoulders, since Italian men devote less time to unpaid family labour than men in any other HETUS (Harmonised European Time Use Surveys) country (Eurostat 2006). For this reason, working wives in Italy experience the so-called "double presence" (Balbo 1978) "second shift" (Hochschild 1989) or "double day" (Ferree 1991) to a larger extent than elsewhere, particularly if they have children. In an institutional context of this kind, female employment may end up by becoming a source of time crunch and emotional strain, reflecting on life satisfaction and marital quality as well as, in turn, on marital stability. Moreover, it may be supposed that in a country as traditional as Italy in terms of family and gender roles and obligations among family members, the presence of a second breadwinner in the home does not strengthen the marriage by contributing to the wealth of the family, but rather undermines the harmony of the marital relationship by taking away the wife from what are considered her primary duties and increasing competition between the spouses. Unfortunately, no empirical evidence is yet available to support or reject this statement.
6.3 Secondly, this study clearly shows that the association between female employment and marital stability seems to be invariant across the different conditions tested here. These findings are not in line with some research carried out in other countries suggesting that these conditions play a significant role in influencing this association. I have not reason to raise doubts about the quality of the dataset used in this analysis in order to explain this discrepancy. The Families and Social Subjects Survey is a large national dataset performed by the Italian National Institute of Statistic part of a wider international survey, that constitutes the empirical base for a considerable amount of studies. Its retrospective sections allow to reconstruct some time-varying variables crucial to the aim of this study, and the large sample size makes it possible to estimate parameters robust enough to analyse the interaction effects of interest. Moreover, several relevant determinants of both marital disruption and wives' employment are stored in the dataset, and consequently can be included in the analysis as controls. Finally, apart from the case of gender ideology already widely tackled, both the key and the control variables do not present serious problems of operazionalization. Thus, the explanation according to which in Italy the association between female employment and marital stability seems to be stable needs to be found elsewhere.
6.4 As regards the constancy of this association across historical periods, it must be considered that women's participation in the Italian workforce has increased at a very slow pace during the last few decades. Growth in the female employment rate began to increase only in the second half of the Nineties and in the first decade of the twenty-first century. For this reason, the cultural and economic changes following the spread of female paid work described in the first section that can impact the association between female employment and marital stability have taken place in very small steps, and some may not have taken place at all. It may be of interest to repeat the analysis put forward here in the coming years, should the female employment rate increase substantially. In addition, work/life balance policies have made little headway in Italy. Consequently, there has been little change in the facilities that such policies provide to help working women face a marital disruption. This fact also contributes to minimizing the changes in the association between female employment and marital stability over the years, since an increase in the disruptive effect of female employment could be expected when more work/life balance facilities are available.
6.5 Focusing on the lack of variation in the association between female employment and marital stability due to the differences in the institutional and cultural contexts of Italian society, it may be maintained that the differences between the North and the South of the country are indubitably strong, but evidently not strong enough to have a real impact on individual behaviours, at least as regards the subject-matter of this study. It is important to underline that even if the female employment rate in Northern Italy is much higher than in Southern Italy, it is not so high in absolute terms, as proved by the fact that in recent years it has always been lower than the average for European Union countries. What can be supposed is that even in Northern Italy female engagement in paid work is still far from being considered as normative, which is the condition that mitigates the disruptive effect of female employment on marital stability. Here, empirical support is provided by data from the European Values Study 1999-2000. This dataset shows that 78.4% of the respondents living in the North of Italy agrees with the statement that a pre-school child is likely to suffer if the mother works. Moreover, 63.8% thinks that a job is all right, but what most women really want is a home and children, and 52.3% supports the contention that for a woman being a housewife is just as fulfilling as working for pay. These figures clearly underline that the attitudes towards female engagement in paid work are such that it still cannot be considered a normative behaviour.
6.6 Finally, turning to the fact that the association between female employment and marital stability does not vary across wives' gender ideology (at least to the extent to which the latter is captured by the experience of a premarital cohabitation), a possible explanation can be found by considering the association between gender ideology and perception of fairness. The first section outlined a theoretical framework according to which female participation in the workforce undermines marriage because of the sense of unfairness that may stem from the division of paid and unpaid work between the spouses, decreasing marital quality and, in the long term, marital stability. Traditional women tend to accept the doctrine of the segregated marital roles, and accordingly may consider that the inequality in this division is not necessarily unfair, while the opposite is true for nontraditional women. Consequently, engagement in paid work is expected to be less a source of marital problems for the former than for the latter. However, the only empirical study carried out in Italy (Carriero 2010) analysing the association between gender ideology and perceived fairness clearly shows that the former has a very small effect on the latter, far from statistical significance. Thus, it can be maintained that the condition implying a variation in the association between female employment and marital stability according to wives' gender ideology seems not to be found in this country. It goes without saying that more scholarly attention on the association between gender ideology and perceived fairness is required in order to support or reject this explanation, partly because the study by Carriero does not discuss the reasons and the implications of the findings in depth.
6.7 A matter of interest is whether other variables not considered in this study might play a role in influencing the association between female employment and marital stability in Italy. An in-depth analysis in this sense is beyond the aim of this study, but in order to shed light at least partially on this possibility, at an explorative level some other possible interactions have been tested here. As already mentioned, neither the rite of wedding nor the marital duration seem to affect the association between female employment and marital stability. Moreover, this association also remains constant across education and number and age of children (table not shown, available on demand). Maybe other scholars could explore more carefully this argument, focusing on the individualization of the interactive variables particularly significative in light of the peculiar Italian institutional and cultural context, and subsequently testing them at the empirical level.
6.8 Eventually, it is necessary to locate the findings of the analysis put forward here in the context of the previous literature on the association between female employment and marital stability, in order to give a contribute to the debate on this matter. This study clearly shows that mixed and controversial findings are found not only analysing the direction and magnitude of this association, but also when attention is focused on possible interactions with other conditions. The previous studies carried out so far provide some theoretical and empirical evidence to the fact that the association between female employment and marital stability tends to vary according to historical period, institutional and cultural context and wives' gender ideology. On the contrary, this study claims that in Italy the association between female employment and marital stability remains invariant across different historical and social conditions. This discrepancy seems mainly due to the peculiar Italian institutional and cultural context, characterised by a relatively high level of stability of the circumstances which should affect the association between female employment and the risk of marital disruption because of their variations. Thus, an indication from this study is that in the research work on this association and the issues that relate to it great caution is required in generalizing the findings out from the institutional and cultural context in which the empirical analysis is carried out, because this context seems to play an important role in these dynamics.
Conclusion7.1 To sum up, on a broad theoretical level this study shows that the specialization and trading model still has some explanatory power for shedding light on the association between female employment and marital stability, at least in a country like Italy which is characterized by strongly traditional family and gender roles. The validity of this model also seems to be maintained across very different contexts and conditions. However, it must be underlined that this study fails to adopt a couple-level perspective, and consequently it is not fully able to take into account the effect of the wider gender system on the perception of female employment, that in turn affects the association between the latter and marital stability. Future research carried out in Italy should continue to shed light on this association, focusing on its variations in the coming years, particularly if there are important changes in the labour market structure, in work-family reconciliation policies or in the incidence of marital dissolution. It would be also imperative to carry out further analysis using datasets containing information on both spouses, in order to consider together both the husband's and the wife's perspective and more in general the gender social structure in which they are embedded. Finally, the effect on marriage of female participation in the workforce should be studied in greater depth, in order to disentangle the direct effect of female employment from the indirect one mediated through wives' earnings. So far this question remains open because of the lack of adequate data.
|Table A1. Transitions from Marriage to Legal Separation. Women-Years Observed in Marriages (Observations) and Legal Separations (Events) in the Final Subsample Used in the Analysis According to Selected Characteristics.|
Notes1On these topics, see the data from Eurostat (epp.eurostat.ec.europa.eu/) and Istat (Italian National Institute of Statistics, http://en.istat.it/).
3The data refer to the year 2008 (2007 for Ireland and Portugal) and come from Eurostat (epp.eurostat.ec.europa.eu/) and Istat (Italian National Institute of Statistics, http://en.istat.it/). Malta is not included in the comparison because to this day divorce is not allowed. At the end of May 2011 a referendum has resulted in a vote in favour of introducing divorce in this country.
7In general terms, an event history can be defined as a longitudinal record of the timing of the occurrence of one or more types of event. In this study, attention is devoted to the marital careers, and consequently the entrance into marriage and the end of the marriage, if it takes place, are the events of interest. Event history analysis is a statistical tool used by scholars to study the duration until the occurrence of an event of interest, where the duration is measured from the time at which an individual becomes exposed to the risk of experiencing this event. This duration is called observation period as well. This kind of analysis is also employed to assess the effect of one or more variables on this duration, or, in other words, on the timing of the events. In this case, I am interested at looking at the effect of the employment status on the duration of the marriage, and its variations according to other variables. A variable that may change status over the course of the observation period is called time-varying, otherwise it is called time-constant (for an overview on longitudinal analysis, see Hosmer & Lemeshow 1999; Menard 2002).
8Regression analysis is a common statistical tool used in research to shed light on the relationships between variables. One of its possible uses is to ascertain the effect of one variable (the main independent variable) upon another (the dependent variable), taking into account as well of other lurking variables (control variables) potentially confounding this effect, since they influence both the dependent and the main independent variables. Aim of the regression analysis is also to assess the statistical significance of the estimated relationships, that is, the degree of confidence that the true relationship in the whole population is close to the estimated relationship in the sample on which the analysis is based (for an overview on regression analysis, see Lewis-Beck 1980).
9In order to give a quantitative dimension to this phenomenon, about the 27% of couples legally separated in 1975 never go as far as divorce. The figure is similar considering couples legally separated in 1980 (Todesco 2008).
10In a regression analysis an interaction effect is found when the effect of an independent variable on the dependent variable differs depending on the level of a third variable (for an overview on the treatment of the interaction effects see Aiken & West 1991).
11It is imperative to recognise that the area of residence is a so-called anticipatory variable, i.e. a variable whose values refer to what is attained by the date of the survey but is used to explain events in life course which occurred before the survey. Accordingly, considering how the area of residence impacts the association between female employment and risk of marital disruption causes a time inconsistency, and the estimate of this effect could be biased. However, in Italy internal mobility has been low in recent decades and mainly confined to short-distance movements (Tomassini et al. 2003), and this is true also in the case of mobility following a marital disruption, confined to movements within the same city, or at most within the same region (Barbagli 1990; Barbagli & Saraceno 1998). Since each area included in the analysis groups various regions, no substantial bias should affect the findings.
12It should be stressed that because of the Italian legislation regulating marital dissolution, some couples can be forced to cohabit before marriage because one or both partners are in the three-year period of compulsory waiting before a divorce decree. However, this fact should not jeopardize the validity of the findings from the subsample used in this study. For married respondents, only first marriages are included in the analysis. For respondents with experience of legal separation or divorce, in order to be an ambiguous case in terms of freedom of choice to cohabit or not, they should be at their second experience of marital disruption and with a previous experience of cohabitation before it. Only a negligible number of respondents has experienced such a marital career, as can be expected in a country with high marital stability.
13In any case, an explorative analysis reveals that the rite of wedding does not affect the association between female employment and marital stability (table not shown, available on demand).
14Education may also be considered a case of anticipatory variable, but this problem is avoided by deleting the very few respondents still in school at the time of marriage formation.
15This figure comes from the difference between the eß (odds ratio) of the category "Not employed" (0.58) and that of the reference category "Employed" (1). The result is statistically significant, since the p-value is well below the conventional threshold (0.05) (for an overview on the interpretation of the coefficients in the logistic regression, see Menard 1995)
16This figure comes from the difference between the eβ (odds ratio) of the category "Not employed" (0.75), referring to the calendar period "1984 and earlier", and that of the reference category "Employed" (1).
17These figures come respectively from the differences between the eβ (odds ratio) of the category "Not employed" (0.75), multiplied by those of the interaction terms referring to the calendar period "1985-1989" (0.79) "1990-1994" (0.57) "1995-1999" (0.82) "2000-2003" (0.72), and the odds ratio of the reference category "Employed" (1).
18This figure comes from the difference between the eβ (odds ratio) of the category "Not employed" (0.60), referring to the area of residence "Northern Italy", and that of the reference category "Employed" (1).
19These figures come respectively from the differences between the eß (odds ratio) of the category "Not employed" (0.60), multiplied by those of the interaction terms referring to the area of residence "Central Italy" (1.08) "Southern Italy" (0.80), and the odds ratio of the reference category employed (1).
20This figure comes from the difference between the eß (odds ratio) of the category "Not employed" (0.58), referring to respondents who did not cohabit before marriage, and that of the reference category "Employed" (1).
21These figure comes from the differences between the eß (odds ratio) of the category "not employed" (0.58), multiplied by that of the interaction term referring to the respondents who did cohabit before marriage (0.96), and the odds ratio of the reference category employed (1).
ReferencesAIKEN, L. S. and WEST, S. G. (1991) Multiple Regression: Testing and Interpreting Interactions, Newbury Park: Sage.
AROSIO, L. (2004) Gli opposti si respingono? Scelte di coppia e stabilità coniugale in Italia, Roma: Aracne.
BALBO, L. (1978) 'La doppia presenza', Inchiesta, 32, pp. 3-6.
BARBAGLI, M. (1990) Provando e riprovando. Matrimonio, famiglia e divorzio in Italia e in altri paesi occidentali, Bologna: Il Mulino.
BARBAGLI, M. and SARACENO, C. (1998) Separarsi in Italia, Bologna: Il Mulino.
BATALOVA, J. A. and COHEN, P. N. (2002) 'Premarital Cohabitation and Housework: Couples in Cross-National Perspective', Journal of Marriage and Family, 64 (3), pp. 743-755. [doi:://dx.doi.org/10.1111/j.1741-3737.2002.00743.x]
BECKER, G. S. (1981) A Treatise on the Family, Cambridge: Harvard University Press.
BOLZENDAHL, C. I. and MYERS, D. J. (2004) 'Feminist Attitudes and Support for Gender Equality: Opinion Change in Women and Men, 1974–1998', Social Forces, 83 (2), pp. 759-790. [doi:://dx.doi.org/10.1353/sof.2005.0005]
BRACHER, M., SANTOW, G. , MORGAN, M., and TRUSSEL, J. (1993) 'Marital Dissolution in Australia: Models and Explanations', Population Studies, 47 (3), pp. 403-425.
BRAMLETT, M. D. and MOSHER, W. D. (2002) 'Cohabitation, Marriage, Divorce, and Remarriage in the United States', Vital and Health Statistics, 23 (22), pp.
BUKODI, E. and RÓBERT, P. (2003) 'Union Disruption in Hungary', International Journal of Sociology, 33 (1), pp. 64-94.
BUMPASS, L. L., MARTIN, T. C., and SWEET, J. A. (1991) 'The Impact of Family Background and Early Marital Factors on Marital Disruption', Journal of Family Issues, 12 (1), pp. 22-42. [doi:://dx.doi.org/10.1177/019251391012001003]
CARRIERO, R. (2010), Il paradosso delle donne "sovraccariche e soddisfatte": risultati da una ricerca empirica su alcune ipotesi esplicative. In AA.VV. (ed.), Giovani Sociologi 2009, Napoli: Scriptaweb.
CHAN, T. W. and HALPIN, B. (2002) 'Union Dissolution in the United Kingdom', International Journal of Sociology, 32 (4), pp. 76-93.
CHERLIN, A. (1992) Marriage, Divorce, Remarriage, Cambridge, MA: Harvard University Press.
CLARKBERG, M., STOLZENBERG, R., and WAITE, L. J. (1995) 'Attitudes, Values, and Entrance into Cohabitation Versus Marital Unions', Social Forces, 74 (2), pp. 609-634.
COOKE, L. P (2006) '"Doing" Gender in Context: Household Bargaining and Risk of Divorce in Germany and the United States', The American Journal of Sociology, 112 (2), pp. 442-472. [doi:://dx.doi.org/10.1086/506417]
COOKE, L. P and GASH, V. (2010) 'Wives' Part-time Employment and Marital Stability in Great Britain, West Germany and the United States', Sociology, 44 (6), pp. 1091-1108. [doi:://dx.doi.org/10.1177/0038038510381605]
COPPOLA, L. and DI CESARE, M. (2008) 'How Fertility and Union Stability Interact in Shaping New Family Patterns in Italy and Spain', Demographic Research, 18 (4), pp. 117-144. [doi:://dx.doi.org/10.4054/DemRes.2008.18.4]
CUNNINGHAM, J. D. and ANTILL, J. K. (1994) 'Cohabitation and Marriage: Retrospective and Predictive Comparisons', Journal of Social and Personal Relationships, 11 (1), pp. 77-93. [doi:://dx.doi.org/10.1177/0265407594111005]
DAVIS, S., GREENSTEIN, T. N., and GERTEISEN MARKS, J. P. (2007) 'Effects of Union Type on Division of Household Labor: Do Cohabiting Men Really Perform More Housework?', Journal of Family Issues, 28 (9), pp. 1246-1272. [doi:://dx.doi.org/10.1177/0192513X07300968]
DAVIS, S. N. (2007) 'Gender Ideology Construction from Adolescence to Young Adulthood', Social Science Research, 36 (3), pp. 1021-1041. [doi:://dx.doi.org/10.1016/j.ssresearch.2006.08.001]
EUROSTAT (2006) 'How is the Time of Women and Men Distributed in Europe?', Statistic in Focus, n° 4, pp.
FERREE, M. M. (1991) 'The Gender Division of Labor in Two-Earner Marriages: Dimensions of Variability and Change', Journal of Family Issues, 12 (2), pp. 158-180. [doi:://dx.doi.org/10.1177/019251391012002002]
FOKKEMA, T. and LIEFBROER, A. C. (2004) 'Employment and Divorce among Dutch Women Born between 1903 and 1937', History of the Family, 9 (4), pp. 425-442. [doi:://dx.doi.org/10.1016/j.hisfam.2004.01.014]
FOX, G. R. and MCBRIDE MURRY, V. (2000) 'Gender and Families: Feminist Perspectives and Family Research', Journal of Marriage and the Family, 62 (4), pp. 1160-1172. [doi:://dx.doi.org/10.1111/j.1741-3737.2000.01160.x]
FRISCO, M. L. and WILLIAMS, K (2003) 'Perceived Housework Equity, Marital Happiness, and Divorce in Dual-Earner Households', Journal of Family Issues, 24 (1), pp. 51-73 [doi:://dx.doi.org/10.1177/0192513X02238520]
GOODE, W. J. (1961), Family Disorganization. In R. K. Merton and R. A. Nisbet (eds.), Contemporary Social Problems. An Introduction to the Sociology of Deviant Behavior and Social Disorganization, New York: Harcourt, Brace & World.
GOODE, W. J. (1962) 'Marital Satisfaction and Instability. A Cross-Cultural Class Analysis of Divorce Rates.', International Social Sciences Journal, 14 (3), pp. 507-526.
GORNICK, J. C. and MEYERS, M. K. (2003) Families that Work: Policies for Reconciling Parenthood and Employment, New York: Russel Sage Foundation.
GREENSTEIN, T. N. (2000) 'Economic Dependence, Gender, and the Division of Labor in the Home: A Replication and Extension.', Journal of Marriage and Family, 62 (2), pp. 322-335. [doi:://dx.doi.org/10.1111/j.1741-3737.2000.00322.x]
GUILLEMARD, A. M. (2005), The Advent of a Flexible Life Course and the Reconfigurations of Welfare. In J. G. Andersen, et al. (eds.), The Changing Face of Welfare: Consequences and Outcomes from a Citizenship Perspective, Bristol: Policy Press.
HÄRKÖNEN, J. and DRONKERS, J. (2006) 'Stability and Change in the Educational Gradient of Divorce. A Comparison of Seventeen Countries', European Sociological Review, 22 (5), pp. 501-517.
HEIMER, C. A. and STAFFEN, L. R. (1995) 'Interdependence and Reintegrative Social Control: Labeling and Reforming "Inappropriate" Parents in Neonatal Intensive Care Units', American Sociological Review, 60 (5), pp. 635-654. [doi:://dx.doi.org/10.2307/2096315]
HIEDEMANN, B., SUHOMLINOVA, O., and O'RAND, A. M. (1998) 'Economic Independence, Economic Status, and Empty Nest in Midlife Marital Disruption', Journal of Marriage and Family, 60 (1), pp. 219-231. [doi:://dx.doi.org/10.2307/353453]
HOCHSCHILD, A (1989) The Second Shift: Working Parents and the Revolution at Home, New York: Avon Books.
HOEM, B. and HOEM, J. M. (1992), The Disruption of Marital and Non-Marital Unions in Contemporary Sweden. In J. Trussell, R. Hankinson, and J. Tilton (eds.), Demographic Applications of Event History Analysis, Oxford: Clarendon Press.
HOSMER, D. W. and LEMESHOW, S. (1999) Applied Survival Analysis. Regression Modeling of Time to Event Data, New York: Wiley & Sons.
JARMAN, J., BLACKBURN, R. M., BROOKS, B. , and DERMOTT, E. (1999) 'Gender Differences at Work: International Variations in Occupational Segregation', Sociological Research Online, 4 (1), pp.
JOHNSON, W. R. and SKINNER, J. (1986) 'Labor Supply and Marital Separation', The American Economic Review, 76 (3), pp. 455-469.
KALLEBERG, A., RESKIN, B. F., and HUDSON, K. (2000) 'Bad Jobs in America: Standard and Nonstandard Employment Relations and Job Quality in the United States', American Sociological Review, 65 (2), pp. 256-278. [doi:://dx.doi.org/10.2307/2657440]
KALMIJN, M. and POORTMAN, A. (2006) 'His or Her Divorce? The Gendered Nature of Divorce and its Determinants', European Sociological Review, 22 (2), pp. 201-214. [doi:://dx.doi.org/10.1093/esr/jci052]
KALMIJN, M., DE GRAAF, P. M., and POORTMAN, A. (2004) 'Interactions between Cultural and Economic Determinants of Divorce in the Netherlands', Journal of Marriage and Family, 66 (1), pp. 75-89. [doi:://dx.doi.org/10.1111/j.1741-3737.2004.00006.x]
KAUFMAN, G. (2000) 'Do Gender Role Attitudes Matter? Family Formation and Dissolution among Traditional and Egalitarian Men and Women', Journal of Family Issues, 21 (1), pp. 128-144. [doi:://dx.doi.org/10.1177/019251300021001006]
KÜNZLER, J. (2002), Paths Towards a Modernisation of Gender Relations, Policies, and Family Building. In F. X. Kaufmann, et al. (eds.), Family Life and Family Policies in Europe (Vol. 2), Oxford: Oxford University Press.
LEE, R. (1983) 'Economic Consequences of Population Size, Structure and Growth', IUSSP Newsletter, 17, pp. 43-59.
LEWIS-BECK, M. (1980) Applied Regression: An Introduction, Beverly Hills: Sage.
LEWIS, J. and OSTNER, I. (1995), Gender and the Evolution of European Social Policy. In S. Leibfried and P. Pierson (eds.), European Social Policy: Between Fragmentation and Integration, Washington: Brookings Institution.
LIEFBROER, A. C. and DOURLEIJN, E. (2006) 'Unmarried Cohabitation and Union Stability: Testing the Role of Diffusion Using Data from 16 European Countries', Demography, 43 (2), pp. 203-221. [doi:://dx.doi.org/10.1353/dem.2006.0018]
LYE, D. and WALDRON, I. (1997) 'Attitudes Toward Cohabitation, Family, and Gender Roles: Relationships to Values and Political Ideology', Sociological Perspectives, 40 (2), pp. 199-225.
LYNGSTAD, T. H. (2004) 'The Impact of Parents' adn Spouses' Education on Divorce rates in Norway', Demographic Research, 10 (5), pp. 122-142.
MAJOR, B. (1987), Gender, Justice, and the Psychology of Entitlement. In P. Shaver and C. Hendrick (eds.), Sex and Gender, Beverly Hills: Sage, 124-148.
MARTIN, K. A. (1998) 'Becoming a Gendered Body: Practices of Preschools', American Sociological Review,, 63 (4), pp. 494-511.
MCGUFFEY, C. S. and RICH, B. L. (1999) 'Playing in the Gender Transgression Zone: Race, Class, and Hegemonic Masculinity in Middle Childhood', Gender and Society, 13 (5), pp. 608-627. [doi:://dx.doi.org/10.1177/089124399013005003]
MCKINNISH, T. G. (2007) 'Sexually Integrated Workplaces and Divorce: Another Form of On-the-Job Search', Journal of Human Resources, 42 (2), pp. 331-352.
MENARD, S. (1995) Applied Logistic Regression Analysis, Thousand Oaks: Sage.
MENARD, S. (2002) Longitudinal Research, Thousand Oaks: Sage.
MILLER, A. and SASSLER, S. (2010) 'Stability and Change in the Division of Labor among Cohabiting Couples', Sociological Forum, 25 (4), pp. 677-702. [doi:://dx.doi.org/10.1111/j.1573-7861.2010.01207.x]
NAZIO, T. and BLOSSFELD, H. P. (2003) 'The diffusion of Cohabitation among Young Women in West Germany, East Germany and Italy', European Journal of Population, 19 (1), pp. 47-82. [doi:://dx.doi.org/10.1023/A:1022192608963]
OECD (2011) Doing Better for Families.
OPPENHEIMER, V. K. (1997) 'Women's Employment and the Gain to Marriage: The Specialization and Trading Model', Annual Review of Sociology, 23 (August), pp. 431-53. [doi:://dx.doi.org/10.1146/annurev.soc.23.1.431]
PAGNINI, D. L. and RINDFUSS, R. R. (1993) 'The Divorce of Marriage and Childbearing: Changing Attitudes and Behavior in the United States', Population and Development Review, 19 (2), pp. 331-347. [doi:://dx.doi.org/10.2307/2938442]
PARSONS, T. (1949), The Social Structure of the Family. In R. Anshen (ed.), The Family: Its Function and Destiny, New York: Harper and Brothers.
PARSONS, T. (1955), The American Family: Its Relations to Personality and the Social Structure. In T. Parsons and R. Bales (eds.), Family Socialization and Interaction Process, New York: Free Press.
POORTMAN, A. and KALMIJN, M. (2002) 'Women's Labour Market Position and Divorce in the Netherlands: Evaluating Economic Interpretations of the Work Effect', European Journal of Population, 18 (2), pp. 175-202. [doi:://dx.doi.org/10.1023/A:1015520411449]
PRESSER, H. B. (2000) 'Nonstandard Work Schedules and Marital Instability', Journal of Marriage and Family, 62 (1), pp. 93-110. [doi:://dx.doi.org/10.1111/j.1741-3737.2000.00093.x]
RAEYMAECKERS, P., DEWILDE, C., SNOECKX, L., and MORTELMANS, D. (2008) 'The Influence of Formal and Informal Support Systems on the Labour Supply of Divorced Mothers', European Societies, 10 (3), pp. 453-477. [doi:://dx.doi.org/10.1080/14616690701833480]
RINDFUSS, R. R. and VANDENHEUVEL, A. (1990) 'Cohabitation: A Precursor to Marriage or an Alternative to Being Single', Population and Development Review, 16 (4), pp. 703-726. [doi:://dx.doi.org/10.2307/1972963]
ROGERS, S. J. (2004) 'Dollars, Dependency, and Divorce: Four Perspectives on the Role of Wives' Income', Journal of Marriage and Family, 66 (1), pp. 59–74. [doi:://dx.doi.org/10.1111/j.1741-3737.2004.00005.x]
ROGERS, S. J. and AMATO, P. R. (2000) 'Have Changes in Gender Relations Affected Marital Quality?', Social Forces, 79 (2), pp. 731-753.
ROSS, H. L. and SAWHILL, I. V. (1975) Time of Transition: The Growth of Families Headed by Women, Urban Institute: Washington.
SALVINI, S. and VIGNOLI, D. (2011) 'Things Change: Women's and Men's Marital Disruption Dynamics in Italy During a Time of Social Transformations', Demographic Research, 24 (5), pp. 145-174. [doi:://dx.doi.org/10.4054/DemRes.2011.24.5]
SASSLER, S. and GOLDSCHEIDER, F. (2004) 'Revisiting Jane Austen's Theory of Marriage Timing: Changes in Union Formation among American Men in Late 20th Century', Journal of Family Issues, 25 (2), pp. 139-166. [doi:://dx.doi.org/10.1177/0192513X03257708]
SASSLER, S. and MILLER, A. J. (2010) 'Waiting to Be Asked: Gender, Power, and Relationship Progression Among Cohabiting Couples', Journal of Family Issues, 20 (10), pp. 1-25.
SAYER, L. and BIANCHI, S. (2000) 'Women's Economic Independence and the Probability of Divorce: A Review and Reexamination', Journal of Family Issues, 21 (7), pp. 906-943. [doi:://dx.doi.org/10.1177/019251300021007005]
SCHOEN, R., ROGERS, S. J. , and AMATO, P. R. (2006) 'Wives' Employment and Spouses' Marital Happiness: Assessing the Direction of Influence Using Longitudinal Couple Data', Journal of Family Issues, 27 (4), pp. 506-528.
SCHOEN, R., ASTONE, N. M., ROTHERT, K., STANDISH, N. J., and KIM, Y. J. (2002) 'Women's Employment, Marital Happiness, and Divorce', Social Forces, 81 (2), pp. 643-662. [doi:://dx.doi.org/10.1353/sof.2003.0019]
SOLERA, C. (2009) Women In and Out of Paid Work: Changes Across Generations in Italy and Great Britain, Bristol: Policy Press.
THOMPSON, R. A. and AMATO, P. A. (eds.) (1999), The Postdivorce Family: Children, Parenting, and Society (Thousand Oaks: Sage).
TJØTTA, S. and VAAGE, K. (2003) 'Union Disruption in Norway', International Journal of Sociology, 33 (1), pp. 40-63.
TODESCO, L. (2008), L'instabilità coniugale: aspetti storici ed evolutivi. In A. Urbano (ed.), Evoluzione e nuove tendenze dell'instabilità coniugale, Roma: Istituto Nazionale di Statistica.
TODESCO, L. (2009) Matrimoni a tempo determinato. L'instabilità coniugale nell'Italia contemporanea, Roma: Carocci.
TOMASSINI, C., WOLF, D. A., and ROSINA, A. (2003) 'Parental Housing Assistance and Parent-Child Proximity in Italy', Journal of Marriage and Family, 65 (3), pp. 700-715. [doi:://dx.doi.org/10.1111/j.1741-3737.2003.00700.x]
TRUSSELL, J., RODRIGUEZ, G., and VAUGHAN, B. (1992), Union Dissolution in Sweden. In J. Trussel, R. Hankinson, and J. Tilton (eds.), Demographic Applications of Event History Analysis, Oxford: Clarendon Press., 38-60.
VAN DAMME, M., KALMIJN, M., and UUNK, W. (2009) 'The Employment of Separated Women in Europe: Individual and Institutional Determinants.', European Sociological Review, 25 (2), pp. 183-197. [doi:://dx.doi.org/10.1093/esr/jcn042]
VOYDANOFF, P. and DONNELLY, B.W. (1999) 'The Intersection of Time in Activities and Perceived Unfairness in Relation to Psychological Distress and Marital Quality', Journal of Marriage and the Family, 61 (3), pp. 739-751. [doi:://dx.doi.org/10.2307/353574]
WAGNER, M. and WEIß, B. (2006) 'On the Variation of Divorce Risks in Europe: Findings from a Meta-analysis of European Longitudinal Studies', European Sociological Review, 22 (5), pp. 483-500. [doi:://dx.doi.org/10.1093/esr/jcl014]
WEISS, Y. (1997), The Formation and Dissolution of Families: Why Marry? Who Marries Whom? And What Happens Upon Divorce. In M. R. Rosenzweig and O. Stark (eds.), Handbook of Population and Family Economics, Amsterdam, New York: Elesevier, 81-123. [doi:://dx.doi.org/10.1016/S1574-003X(97)80020-7]