Union Stability in Mexico:
Cohort Change in Educational Differentials Abril Arteaga
Master’s Thesis in Demography
Multidisciplinary Master’s Programme in Demography, Spring term 2012 Demography Unit, Department of Sociology, Stockholm University
Supervisor: Elizabeth Thomson
UNION STABILITY IN MEXICO:
COHORT CHANGE IN EDUCATIONAL DIFFERENTIALS ABRIL ARTEAGA July 22, 2012 Introduction
Western Europe and North America have experienced changes in partner structures and dynamics throughout the 20th century. Marriage has become less common, divorce rates have increased and in general less stable forms of unions are emerging (Hobcraft and Kiernan, 1995; Blossfeld and Huinink, 1995). Increases in divorce and in cohabitation are the two most prominent features of changing family behaviour in Western society. Furthermore, cohabitating couples are more likely to dissolve their union than are married couples. Although high proportions of consensual unions are formalized by marriage, increases in cohabitation and higher rates of separation than in marriage produce very high levels of union dissolution.
Both cohabitation and divorce are associated with education. Educational differentials appear, however, to shift from positive to negative as union dissolution becomes more common in a society. When cohabitation and divorce are uncommon, higher education is related to the de-‐institutionalization of marriage and alternative family practices. As they happen to be more universal and acceptable, it is the disadvantaged who are more likely to cohabit or divorce (Dronkers, Kalmijn & Wagner 2006).
In this paper, I examine the effects of female education on union instability in Mexico and the role of cohabitation. Although consensual unions have long been customary in Mexico, their nature may have changed. Both cohabitation and divorce have increased in the past decades. I test whether the educational gradient in union dissolution is different for unions formed when divorce and cohabitation were less common than at later periods.
Divorce, Cohabitation and Separation
In the last fifty years, there has been a continuous transformation in union patterns of societies around the world. In most industrialized countries, marriage has become a weaker institution (Hobcraft and Kiernan, 1995; Huinink, 1995) increasing the odds of individuals to experience separation or divorce. Traditional living arrangements such as marriage are considerably less attractive than around the mid-‐1960s when almost everyone entered into it. The so-‐called ‘golden age of marriage” (Festy 1980) has been left in the past in the last couple of decades; some scholars have even predicted the family’s demise (Pinnelli Hoffmann-‐Nowotny and Pax, 2001). With variation in timing and speed, most industrialized countries have been facing an increase in union instability.
Figure 1 presents crude divorce rates (number of divorces per 1000 population) in 1970 and 2006 for different countries in Europe and North America. Unions became less stable, as reflected by the increase in the crude divorce rate over the observed period. When considering the increase in the divorce rate it should be noted that national laws did not allow divorce in several countries until recent decades or it also may be, in part, due to divorces occurring in countries where separation was not previously recorded.
Figure 1 shows that in 2008 the crude divorce rate was highest in United States (almost 4 per 1 000 inhabitants) and Belgium (3,2). The lowest crude divorce rates were recorded in Mexico (0.7 per 1 000 inhabitants in 2010). A number of other catholic Member States also recorded relatively low crude divorce rates, including Poland and Italy. The largest increases in crude divorce rates were recorded in Spain because it was not legal until 1980.
FIGURE 1 DIVORCE RATES 1970-‐2008
Source: Eurostat (2010) and United Nations Statistical Division (2010).
Note: * Data refers to 2007 for the United Kingdom, Australia, Canada, Japan, Greece, Ireland and Mexico; 2006 for the United States, France, Israel and Chile.
Cohabitation
Cohabitation is the other most important transformation in partnership arrangements of the 20th century (Kiernan 2002). At the beginning of the 1960´s, marriage was a prerequisite for union formation; 40 years later in Northern and Western Europe, cohabitation has become an accepted alternative to marriage (Hoem 1995, Toulemon 1997, Liefbroer 2003). Cohabitation is important for union stability because individuals who cohabit outside of marriage and people cohabiting prior to marriage have a higher risk of union dissolution than couples entering directly into marriage (Dourleijn & Liebfbroer 2003).
The proportion of cohabiting couples in European countries has increased during the last decades but research shows that incidence and patterns vary across nations. For example, Kiernan has found that cohabitation has high proportion in the Nordic countries (Denmark, Sweden, Finland) and also in France, in contrast with southern European countries and Ireland, where only a small minority of the population is cohabiting. Some of the industrialized countries that have been distinguished by Kiernan for having an average level of cohabitation among European countries are Netherlands, Belgium, Britain, Germany and Austria. The fraction of cohabitors among current co-‐resident couples is nowadays around 12% or above in many European countries like Denmark, Finland, France, the Netherlands, ,Belgium and Sweden (Figure 2) andi it has been raising for most of the countries presented in the graph for the last 10 years.
FIGURE 2 COHABITING COUPLES AS A FRACTION OF ALL CURRENT CO-‐RESIDENT COUPLES IN DIFFERENT COUNTRIES (2000-‐2010)
Source: Gallup World Poll for 2010 data and OECD Family Database for 2000 data.
Prinz (1995) identifies four progressive stages in the diffusion of non-‐marital cohabitation in Europe: 1) Cohabitation as an avant-‐garde lifestyle, 2) as a preliminary stage before marriage, 3) as a socially accepted living arrangement (even when there are children) and finally 4) as a partnership equal to marriage. Direct marriage in Nordic countries was already uncommon among the cohorts born in 1950´s and today less than 5% marry without cohabiting (Population & Societies, 2006). Living together before marriage is virtually universal in Sweden: by the late 1970s, 96 % of women who had married had cohabited first. And only 20% of the cohabiting couples married within three years of starting to live together. In France by the 1990s, marriage was still the dominant form for couples living together though cohabitation is now widespread. In
0 5 10 15 20 25 AUSTRIA
BELGIUM CZECH REPUBLIC DENMARK FINLAND FRANCE GERMANY IRELAND ITALY NETHERLANDS POLAND PORTUGAL SPAIN SWEDEN UNITED KINGDOM UNITED STATES COHABITATION 2000 COHABITATION 2010
1965, 10% of new couples cohabited before marriage, while thirty years later as many as 90% do (Rydell 2002).
In contrast, Latin countries remain in the first stage, being a precursor in the transition to marriage. In Spain and Italy, more than 90% of women marry directly. According to the Fertility and Family Surveys, only 11% of Spanish women born in 1960-‐1964 entered cohabitation as a first union, and 7% did it in Italy (Domínguez, Castro Martín, and Mencarini 2007). In southern European countries, the diffusion of cohabitation has been rather hesitant; indeed, this type of conjugal arrangement was either very rare or non-‐existent in the first half of the century.
Nevertheless, referring to the theory on the Second Demographic Transition, Van de Kaa (1987, 2001, 2004) and Lesthaeghe (1991) argue that, in the coming decades, a spread of non-‐marital unions will occur, even within the Mediterranean area. Van de Kaa states that, as societies develop and people’s cultural representations change, “a second demographic transition will inevitably follow” (Van de Kaa 2001, p. 325).
Cohabitation and Union Instability
Cohabitation less often includes childbearing than marriage; marriages tend to present a higher degree of commitment. The presence of younger and fewer children in a union negatively affects the risk of dissolution in married couples (Waite and Lillard 1991). Cohabiting couples with children appear still to be less stable than married couples with children, a finding that holds in the USA (Carlson et al 2004; Manning et al 2004; Wu et al 2001), Canada (Le Bourdais et al. 2000), and Europe ( Heuveline et al. 2003; Kiernan 2002).
Cohabiters and married couples who cohabit previous to marriage have higher risk of union dissolution than those who enter directly to marriage (Liefbroer and Dourleijn 2006). Most studies suggest that such instability comes from cohabiters’ characteristics. Cohabitors hold less conventional values and attitudes than those who marry directly (Axinn and Thornton 1992; Balakrishnan et al. 1987; Berrington and Diamond 1999; DeMaris and MacDonald 1993; DeMaris and Rao 1992; Lillard et al. 1995), They express weaker commitment to marriage in general or hold higher expectations about the quality of unions (Bennett et al. 1988; DeMaris and Rao 1992; Lillard et al. 1995; Teachman et al. 1991; Thomson and Colella 1992). Finally, cohabitors possess socioeconomic or personality characteristics that are linked to an increased risk of union dissolution (Berrington and Diamond 1999; Hall and Zhao 1995.
However, a recent study from Kiernan 2002 suggests that increased risk of union dissolution among former cohabiters might not be as universal as is often assumed. Comparing union dissolution rates for nine European countries, Kiernan reported that in five of them, prior cohabitation did not produce higher union dissolution rates than direct marriage. At the same time, Kiernan identifies significant differences in the excess risk of union dissolution for current cohabiters compared with women who married without prior cohabitation, ranging from about a 50% excess risk in former East Germany to almost a fivefold excess risk in Norway and Switzerland.
Education and Union Patterns
According to the (SDT) Second Demographic Transition, alternative living arrangements were a consequence of a more knowledgeable society. Thus, as populations become wealthier and more educated, they tend to shift from survival to post-‐materialism values. The new living arrangements documented above are thought to represent in part the expression of a more secular and antiauthoritarian society (Van de Kaa 2001). The SDT assumes that typical demographic signs in union arrangements (rise in divorce, separation and cohabitation) appear in those societies that equally develop in the direction of capitalist economies, with multi-‐level democratic institutions, and greater accentuation of Maslowian “higher order needs” (Lesthaeghe 2010). Some of the trends pushing the spread of cohabitation are, for example an increasing acceptance of expressing sexuality, a rapid weakening of social control by institutions, increased female control over reproduction, a rise in the importance placed on the quality of the adult dyad, the development of more equal intra-‐union patterns of exchange and the discovery of opportunity costs by women (Schröder 2005).
In some of the analysed countries cohabitation was, if not common, at least regularly practiced in certain areas or by certain groups in earlier periods. Northern and western countries have increased their cohabitation rates across the time and spread this type of union in the society. On the other hand, the UK started with cohabitation patterns already in the 2nd half of the nineteenth century with the industrial working classes (Hasley and Kiernan, 1987) when cohabitation was a popular or more frequent living arrangement just within the less educated segments of the population. Then it was slowly spread in the society, until the mid-‐1990’s when an increase of 53% in portion of cohabiting couples among pairs was experienced. The UK is catching up with the Nordic and Western cohabitation rates.
Even when this modern living arrangement has become relatively widespread in most European countries, it has been relatively uncertain in Spain and Italy, considered as latecomers in this transition. In Spain, cohabitation was slowly started by the late 1960s when mainly students began to cohabit on a large scale. Cohabitation then spread to other groups during the 1970s. Before 1977 however, cohabitation was still marginal, but it is now more common in all social groups. In Italy, if not common, cohabitation is at least more currently practiced in the northern regions of the country, where the slow diffusion of cohabitation has been more pronounced. The percentage of non–marital unions continues to be extremely low in the south. In the whole of Italy, cohabitation tends to be a precursor of the transition to marriage: most couples cohabit as a pre-‐ stage towards a conjugal union, and do so especially when giving a birth (Löffler 2009).
In short, the North-‐South divide in cohabitation prevalence has attracted considerable attention in the demographic literature. But there are diverging interpretations. In some studies southern European countries the diffusion of cohabitation is considered confined to a highly selective population with little prospects to spread to a larger population (Nazio and Blossfeld, 2003). Nevertheless, this viewpoint does not fit with the relatively large increase in non marital births reaching 26.8% in Spain and 17.3% in
Italy in 2005. In other studies the low prevalence of cohabitation is interpreted as a delay in the adoption of innovation behavior. The increase of cohabitation, according to Van de Kaa is “inevitable” or just a matter of time, and Southern European countries are portrayed as later-‐comers in the North-‐Western European pattern of cohabitation (Van de Kaa 1987).
There is considerable debate and conflicting evidence about the effect of education on union disruption. Becker’s theory posits that educational level is a key variable to increases autonomy and provides the base of economic independence to individuals, letting them having more freedom to decide whether to enter or exit a union. Education influences the usefulness of getting married or staying in a union merely for financial reasons. Furthermore, social and economic factors like economic wealth, values and legislation influencing the cost of divorce and benefits of marriage have to be considered in union separation.
Another mechanism is the influence of values on willingness to divorce; more educated individuals have higher possibilities to develop more liberal values (Levinger, 1976) and therefore greater willingness to leave an unhappy relationship. Divorce legislation is an obvious factor in marriage dissolution (Härkönen & Dronkers 2006). It has been found that, at least in the short term, the effect of liberalization of legislation in divorce laws is to increase divorce rates (Friederberg 1998; Wolfers, 2003). On the other hand, Goode (1970: 85-‐86) claims that the strictness of divorce has different effects depending on the socioeconomic level of the individuals. The lower classes benefited least from strict rules, being less resourceful in finding ways to get around them. With more liberal laws, “the greater difficulties of lower class family life were permitted an expression in divorce” (Goode, 1970: 85-‐85, as cited in Härkönen & Dronkers 2006: 502-‐503).
On the other hand, there is the theory of the encouraging role of education in “developing lasting relations”. Education helps individuals to grow socially and culturally, and to develop economic and cognitive skills (Amato, 1996; Hoem, 1997; Ono 1998; Dronkers, 2002). Such characteristics are useful to overcome marital problems and promote satisfactory partnerships. If education is considered as a means of improving resources, it can be implied that it also allows couples to have a better economic and more stable relationships. Furthermore, as women’s economic resources increase, the more important is her participation in the status of the whole family (Oppenheimer, 1997) and less likely the man would like to experience divorce. Moreover, if divorce is relatively rare in a society, such behaviour is more innovative, and therefore more likely among the well educated with more liberal values.
Hoem (1987) found a shift in Sweden where the relationship between female education and divorce has changed from positive to negative; similar shifts have been found by Chan and Halpin (2005) in the United Kingdom. Härkönen and Dronkers (2008) examine the relationship between female education and the risk of divorce over time in 16 European countries and the United States. They found that women with higher education had a higher risk of divorce in France, Greece, Italy, Poland, and Spain. In Austria, Lithuania and the United States, the educational gradient of divorce is negative.
Furthermore, the educational gradient becomes increasingly negative in Flanders, Finland, France, Greece, Hungary, Italy, Lithuania, Poland, Sweden, and the United States. The de-‐institutionalization of marriage and unconventional family practices in these countries seem to be linked with a negative educational gradient of divorce, while welfare state expenditure with a more positive gradient.
The Mexican Case
Mexico provides another case study for analysing the educational gradient of divorce. Mexico has experienced an extremely rapid industrial development characterized by the explosive growth of urban areas, especially after the 1960´s. The hurried expansion of the country helped its economic development, but some social consequences (e.g. poverty, social differences, and education concerns) have been attributed to the industrialization and transformation of Mexico in such a short period. Both educational systems and family behaviours have changed in the context of these broader economic changes.
Education in Mexico
The educational model was used as a way of preparing the population to experience the economic changes that the country was living. First, teaching methods were reformed in order to stimulate and support the modernization process in which the Mexico was embedded at the end of 1980´s and beginning of the 1990´s. However, the NAFTA agreement in 1994 changed governmental priorities and the education model of Mexico was updated once more. There was a shift in education from a nationalistic system to a system relying on values related to hard work, competitiveness, and productivity and international literacy.
In other words, in the 1970´s Mexico was characterized by having an authoritarian and centralized form of government, whereas as time passed, the country adopted a new socio-‐economic model alike to the United States and Canada. The main difference in models was that the traditional one (beginning of the 1970´s) had a more family oriented society, with strong gender roles, pro-‐nuptial culture and religion as the base of the values and attitudes of the population, whereas the modern and new model was encouraging values related with freedom, individualism, democracy and being a more self-‐expressive society.
Moreover, the role of the woman was very important in this socio-‐economical transformation. Thus, between 1970 and 2000 women’s access to education and employment improved considerably in the country. Among 20 to 24 year-‐olds, the proportion of women with secondary education climbed from 3% to 38% and the rate of female labour force participation more than doubled from 21% to 44% (INEGI, 1997, 2001). Those educational transformations might have contributed to shape union stability in Mexico by giving woman more liberal values. From being a less educated society with union dissolution as an uncommon event and strict norms against divorce, Mexico has been moving towards a more educated society with a wider spread acceptance of partnership dissolution.
Cohabitation in Mexico
The Mexican experience might, however, differ from other cases because cohabitation has been more common in the context of low divorce rates, unlike other western European countries where such phenomena have been acting in tandem (e.g. Sweden, France, Spain). In Latin America, consensual unions have been prevalent since the Spanish Colonization, as either an alternative or precursor to marriage. Couples formed by Spanish male colonizers and indigenous women in the 16th and 17th centuries can be considered as their historical antecedent (Brenes Camacho 2008). Many Spanish colonizers where already married in their native Spain and therefore when they came to America could not formalize their unions with the American Aborigines. This behaviour has been repeated across the time, probably for the same reasons, as a way of having more than one union when divorce was not an option. Nonetheless, it is important to highlight that although cohabitation across the time it has been historically more frequent in Mexico than in other industrialized countries, religious or civil marriage is still the most frequent and socially recognized way of starting a union in Mexico (Brenes Camacho 2008). Marriage has not been replaced by cohabitation as the initial stage of family formation and certainly it has not followed the same development of the northern and western European countries since the mid 1960s, where the rise of cohabitation has been followed by declining marriage rates and increasing divorce rates.
Because consensual unions in Mexico are more as a cultural behaviour than an expression of liberal values, Mexico could follow a more southern European pattern, where the family model is characterized by widespread multigenerational living arrangements, high marriage rates, strong kinship networks, low female employment, low fertility and strong family-‐oriented values.
Figure 3 shows the dramatic rise in divorce that has occurred in the last years in Mexico, similar to the Western European countries. Divorce usually becomes more common when societies become more industrialized and Mexico has not been the exception. However, it is important to mention that a common practice in Mexico is separation without divorce, so that divorce rates underestimate union instability (Ojeda & Gonzalez 1994). Marriage dissolution has occurred more than two centuries ago in Mexico and has become accepted in the society. Nevertheless, there are some religious ideas that question such practice.
FIGURE 3 NUMBER OF REGISTERED DIVORCES PER 100 REGISTERED MARRIAGES IN MEXICO (1980-‐2009).
Source: Eurostat Source Data from INEGI 2010 ( National Institute for Statisticis and Geography in Mexico) Note: * The recognition of divorce has represented a total change in the attitudes towards family dynamics in the country given the religious principles of Catholic marriage as a definitive step in life.
There is not conclusive information showing a clear relation between the socioeconomic differences of the population and divorce patterns. On one hand, some studies establish a negative relation between the socioeconomic position of the families and the marital dissolution (Garcia & Rojas 2002). Other investigations have revealed that in the presence of a modern, urbanized and more gender equal society, divorce rates are higher. Thus, it has been proposed that in Mexico the propensity of marriage dissolution has increased in the more educated sectors of the society where the women have access to education and most likely to the labor market. Furthermore, legal divorce demands economic costs that the lower classes are not able to afford.
In recent years, the risk of first union dissolution (marriage and cohabitation) is greater for couples entering into a union at younger ages, highly educated and living in an area with higher living standards area. Cohabiting couples are most likely to separate, followed by legally married couples, and those with legal and religious marriages (Ojeda & Gonzalez 1994). The rise in divorce appear to be more related, as an expression of a non-‐conformist attitude, to protest against authority, a way of manifesting one’s own freedom against conventions whereas the prototype of cohabitation appears to be the opposite. Thus, where divorce is more common among couples that are educated, younger, more secularized, autonomous, cohabitation appears to be coming from an old and traditional model inherited from the colonial times and present in the less educated segment of the population. Such social differences reflect social inequalities in the civil status of the individuals and probably it points out or reflects the educational differences among social groups existing in a country of contrasts like Mexico. Education inequalities reflect unstable differences in marital life across time and space in the (mainly negative) consequences of divorce; however the effect in cohabitation might differ given the social context of the country.
Summary
As a result of the changes that have occurred in Mexico, the relationship between education and divorce or, more generally, union dissolution, may have changed. Better educated young cohorts with an egalitarian world view and greater emphasis on higher order needs (i.e. self-‐actualization, expressive values, recognition) are those who may be most likely to initiate the new behaviours, but as union dissolution becomes more common, the resources that keep unions together become more differentiated by education than values.
Hypothesis: In the 1970s, women with a higher level of education have a higher risk of experiencing separation than women with a lower educational level. Among later union cohorts, the effect is reversed.
Data and Analyses
The analysis is based on data from ENADID 97, a national household survey conducted in 1997 and considering 88,802 women aged 15-‐54 and covering the following topics: Household characteristics, number of members of the household, general data, migration, education, civil status and economic characteristics, mortality, fecundity, number and status of births, contraception and reproductive information and union history. From the full sample, 55,877 had entered a first union at age 13 or older and are included in the analysis.
The risk of dissolution of the first union was analysed, whether consensual or marital, in relation to education and union status at the start (cohabiting, married). Respondents were asked their highest level of education, choosing from the following: none, primary, junior high, high school, teacher education, university degree or more. I combined those with no education and only primary education as very few women had no education. A substantial number of women did not know their education or did not respond and are included in a separate ‘unknown’ category. Control variables include:
• Age at first union: Continuous and fixed variable indicating the age of interviewed women in years when their first union took place.
•Year of Union: Fixed variable compiling the year of first union combined into different categories: 1950-‐1969, 1970-‐1979, 1980-‐89, 1990-‐1999. The first two decades were combined to produce enough cases for the interaction between education and calendar time.
•Birthplace Development: Respondents reported their place of living at the time of the interview. Characteristics of the place were to generate a score on the Human Development Index, categorized (in the original data) as low, medium and high. The creation of a single statistic or score serve as a frame for measuring both social and economic development of every state and its calculated taking into account the following areas: Education, life expectancy, income and health.
•Religion: Respondents were asked their religious denomination. Possible responses were: Catholic, Protestant, Jewish, other or none. Because the large majority responded ‘Catholic’ religious affiliation is coded as Catholic or not.
A Cox proportional hazards model was applied to estimate the risk of union dissolution among the interviewed women. The survival time was defined by the length in years between union and separation or censoring at the interview.
RESULTS
Table 1 contains information about the research population that we are analysing. Most of our target individuals are women whose first union has not disolved, and most were married rather than cohabiting at first union. The women are mostly catholic, and living in areas considered to have better living standards. Higher proportions of women had entered into a union between 1980 and 1990 rather than earlier, and most have only elementary or junior high education.
TABLE 1 DESCRIPTIVE STATISTICS FOR WOMEN AT UNION FORMATION
Figure 4 presents the Kaplan-‐Meier plot where all observations and censoring for the analysis are independent and shows the base line survival curve (relationship is intact). The probability decreases steadily with the time in union, with a minimum above 80%.
FIGURE 4 KAPLAN MEIER PLOT FOR SURVIVAL ANALYSIS
Table 2 provides estimates from the additive and interaction models. In the first column are estimates for the additive multivariate model, and the second for the full interaction model. As the first column shows, women with high or medium levels of education have higher risks of union dissolution than poorly educated woman. The exception is the group of women with teacher education. For this medium highly educated group, the divorce or separation risk is lower than for the poorly educated. The positive estimates in our model resemble the pattern presented in countries such as France, Greece, Italy and Spain.
TABLE 2 ADDITIVE AND INTERACTION MODEL
There is also, as expected, a higher risk of separation across all cohorts. The risk of first union dissolution is greater for those individuals whom enter into a union at younger ages, and for those living in an more developed area. The negative coefficient on the hazard model means that Catholics are less likely to separate. Thus, as is the case in some other religious countries, Catholics are less likely than Protestants or non-‐ religious persons to separate.
The difference between the chi-‐square statistics for the additive and interactive models is 1812-‐1776 = 36 and the difference in the degrees of freedom is 28-‐13=15. Chi-‐square = 36 with df = 15 is statistically significant at the .001 level (30.58). This means that the educational differences are not the same across union cohorts.
Changes in educational differentials can be seen more clearly in Figure 4. The bars represent the relative risk ratios (excluding the unknown category to simplify the presentation) for education within cohort. It shows that university-‐educated women who formed unions in the 1950s and 1960s had lower rates of separation than those with junior high or high school (but not lower than teachers or those with less than junior high). But the next union cohort, 1970s, produced relatively high rates of separation for the university women. But for those forming unions in the 1980s, university education was not different from other education, except the continuing lower rate for women with less than junior high. The newest unions, those formed in the 1990s, had similar separation rates for all education groups, event the least well educated were catching up.
FIGURE 5 SEPARATION RISKS
Conclusion and Discussion:
In this article, we have examined the relationship between female educational attainment and the risk of divorce across cohorts entering into first unions during different periods (1940s, 1950s, 1960s, 1970s, 1980s, and 1990s) with data from the Mexican ENADID 97. Beginning with an hypothesis by William J. Goode (1962), we expected educational differences to vary across cohorts such that the effect of education on divorce would become more negative over time.
Cox regression models showed that, overall, the educational gradient of divorce was positive in the Mexican case, it follows the patterns presented in Catholic countries such as France, Greece, Italy, Poland and Spain. The results also seem to support Goode´s hypothesis based on a positive association between cost of divorce and its educational gradient; “Lower female educational attainment increases the gains from marriage, and decreases the risk of divorce”. In other words, for the least educated, the benefits of the union are higher than the benefits of divorce or separation. In contrast, looking at the most educated sector (University level), Goode’s hypothesis appears to be not fully able to explain the patterns. The university group entering into union between 1940 and 1960 had lower rates than those with less education such as junior high or high school, but not lower than teachers or those with less than junior high education. This pattern could be due to continuing strong social, and economic barriers for such groups of the society. For that reason, the effect of female education on separation might depend on the net effect of some other individual and contextual characteristics.
Between the 1940s until the end of the 1960s, the Mexican government had as a goal accelerating the transition of Mexico from rural to industrialized country. Some of the actions to speed up the process included growth in the number of private and public schools across the whole country and spreading out education among all the population regardless of geographic location, social class or gender. Women were encouraged to attend higher levels of education, as the result of economic pressure, yet not as the result of an ideological change. According to the National University (UNAM) Statistics, the percentage of women among all those enrolled in the institution was: 20.73% for 1940, 18.26% for 1950 and 17.62% for 1960. Percentagewise, the total population receiving higher education at the National University (UNAM) during the 1950s and 1960s was 0.06% and just the fifth part of this percentage was formed by women (0.012%). Female participation in the education system was not substantial, so even when such a group of women were moving towards emancipation and acquisition of more liberal views, the whole system was not ready to accept divorce or separation as a civil status.
In contrast, the group of women entering into a union in the 1970s and receiving higher education at almost the same period, were part of student movements in favor of
changing the Mexican social and economic system. Peculiarly, such a group with more liberal values appears to have the strongest positive influence between education and risk of separation from all the cohorts. The late sixties and the seventies were decades when young female students joined social movements, following their ideals, and
education was a priority in their life to achieve their principles. One can say that, from this period Mexican women started to defend their interest and express opinions. Gisela Espinosa (University Director and Researcher, Program of Gender studies at UNAM) points out that student movements (1965-‐1970) were a “benchmark” in the creation of a new society, where Mexican women converted their status from being passive spectators to be active key players in the political social and economic development of the country.
At the end of the seventies, education was spread among all social sectors of the population and gender differences decreased slowly. From 1980 to 1989, the average percentage of the student population composed of women in universities and
technological institutes was 34.38%. Although in Mexico, the educational arena has historically been strongly dominated by men, the presence of women in not just the university sector, but their increase in all education grades, became more common place.
Such transformation in the educational arrangements of Mexican women can be linked to observed patterns within the newest unions, those formed in the 1990. Mexican women experiencing first union in the 1990´s had similar separation rates for all the education groups, even the least well educated were catching up with all the other groups. From this decade, it is perceived an increase in the number of middle class women and lower classes that are looking to express their points of views and disagree with the societal norms. A new society with more educated women (not necessarily at university level but in all the educational levels) who enter into unions in the 1990’s started building their own identity and considering new living arrangement in which union disruption was an option.
The role of education across time in union disruption has been played with caution, thus until recent years when schooling has been widely spread among all sectors across the country, divorce and separations rates have started to present similar patterns across all educational levels (from basic to university or higher). We interpret our results in tandem with the understanding women’s behavior towards separation or divorce is clearly influenced not only by characteristics of the involved individuals, but also by the characteristics of their societies and their time.
Some researchers consider divorce to be a result of growing individualization and secularization in a society. Mexico has been experiencing such transition in the last 50 years and trying to accomplish such process with the support of women’s education and increasing education has put pressure on challenging traditional values of marriage and raising children, leading to an increased divorce rate.
The increase in divorce is one of the most visible changes in Mexico and despite this general trend, divorce is still more common among certain social groups. Cohorts with less secularization and individualization present lower divorce rates. If a higher educational level of women produces a higher level of individualization, there should be a positive relation between educational level of individuals and their divorce risk, as it can be observed in Mexico.
However, it is possible that educational gradient becomes more negative with the pass of the time and at some point in Mexico, higher education levels can be related to lower divorce rates. This reverse pattern is due in part to the tendency for more highly educated individuals to marry later and live in areas with higher living standards, has allowed for a more stable social structure. Through education and secularism in societies can be impulses for change, the choice of marriage between two people rapidly becomes a practical decision rather than one made accomplishing a prerequisite or in other words what “should be” expected from the individuals’ life arrangements according to a traditional society.
BIBLIOGRAPHY
1. Amato, Paul R. 1996. “Explaining the intergenerational transmission of divorce”. Journal of Marriage and the Family 58, 3, 628–640.
2. Becker, Gary S, Landes, Elisabeth M and Michael, Robert T. 1977. "An Economic Analysis of Marital Instability," Journal of Political Economy, University of Chicago Press, vol. 85(6), pages 1141-‐87, December.
3. Becker, Gary S and Tomes, Nigel. 1976. "Child Endowments and the Quantity and Quality of Children," Journal of Political Economy, University of Chicago Press, vol. 84(4), pages S143-‐62, August.
4. Bennet N. A., Blanc A., Bloom D.1988. ”Commitment and the modern union: assessing the link between premarital cohabitation and subsequent marital stability”, American sociological review 53: 127-‐38.
5. Balakrishnan, T.R., K.V. Rao, E. Lapierre-‐Adamcyk and K.J. Krotki. 1987. "A hazard model analysis of the covariates of marriage dissolution in Canada." Demography, Vol. 24, No. 3, pp. 395-‐406.
6. Berrington, Ann and Diamond, Ian. 1999.” Marital dissolution among the 1958 British birth cohort: The role of cohabitation”. Population Studies, 53, (1), 19-‐38. (doi:10.1080/00324720308066)
7.. Brenes Camacho, Gilbert. 2008.” Economic well-‐being among elderly couples in marriage and cohabitation in Mexico”. Población y Salud en Mesoamérica.
8. .Blossfeld, H.-‐P. and J. Huinik. 1991. “Human Capital Investments or Norms of Role Transition. How Women’s Schooling and Career Affect the Process of Family Formation” American Journal of Sociology 97(1): 143–68.
9. Carlson Marcia McLanahan, Sara and England Paula. 2004. “Union Formation in fragile Families” Demography. Author manuscript; available in PMC 2011 September 8.Published in final edited form as: Demography. 2004 May; 41(2): 237–261.
10. DeMaris, Alfred and Rao. 1992. “Premarital cohabitation and subsequent marital stability in the United States: A reassessment”. Journal of Marriage and the Family, 54, 178-‐190.
11. DeMaris, Alfred and MacDonald, William. 1993. “Premarital Cohabitation and Marital Instability: A Test of the Unconventionality HypothesisJournal of Marriage and Family”. Vol. 55, No. 2 (May, 1993), pp. 399-‐407Published by: National Council on Family Relations, Article Stable URL:http://www.jstor.org/stable/352810
