UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Do resources matter? : the cross cultural effect of gender ideology on the relationship between resources.. Gozjolko, Kristi Lynne 2006

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2006-0472.pdf [ 3.12MB ]
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0092699.json
JSON-LD: 1.0092699+ld.json
RDF/XML (Pretty): 1.0092699.xml
RDF/JSON: 1.0092699+rdf.json
Turtle: 1.0092699+rdf-turtle.txt
N-Triples: 1.0092699+rdf-ntriples.txt
Original Record: 1.0092699 +original-record.json
Full Text
1.0092699.txt
Citation
1.0092699.ris

Full Text

DO RESOURCES MATTER? THE CROSS CULTURAL EFFECT OF GENDER IDEOLOGY ON THE RELATIONSHIP BETWEEN RESOURCES AND MARITAL POWER by Kristi Lynne Gozjolko B.A., University of British Columbia, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in The Faculty of Graduate Studies (Family Studies) UNIVERSITY OF BRITISH COLUMBIA June 2006 © Kristi Lynne Gozjolko, 2006 Abstract In most investigations of marital power the focus of the study is on an individual's characteristics within the relationship with little consideration of the effect of the social context in which the relationship exists. This investigation not only tests the individual's characteristics but also uses country context as a moderator. This study investigates the relationship between resources and marital power and the moderating effect of gender ideology as the country context. Using the 2002 International Social Survey Programme (ISSP) Family and Changing Gender Roles Module III data set allows a cross cultural comparative investigation to occur. From the ISSP 2002 one country is chosen to represent each context, egalitarian (Sweden), traditional (Philippines) and transitional (United States). Using multinomial logistic regression it is found that the resources of relative education and relative income do not directly increase the participant's power but rather, depending on the gender of the respondent, resource contribution is found to increase or decrease the partner's power. Country context did not moderate the relationship between relative resources and marital power as expected. This investigation does not necessarily suggest that cultural context does not matter to marital power but simply indicates that more research is needed. Future cross national comparative research could include more countries in the analysis to have a greater understanding of the relationship between resources and marital power across countries. ii Table of Contents Abstract ii Table of Contents iiList of Tables v List of Figures v Acknowledgements vi Dedication viIntroduction 1 Literature Review 2 Hypothesis 16 Methods 8 Analysis 30 Results 1 Sample Description 3Logistic Regression Analysis 35 Discussion 50 References 62 Appendix 9 iii List of Tables Table 1. Sampling and Data Collection Procedures 22 Table 2. Descriptive Statistics of Research and Control Variables 32 Table 3. Male and Female Participants Multinomial Logistic Regression 37 Table 4. Male Participants Multinomial Logistic Regression 41 Table 5. Female Participants Multinomial Logistic Regression 45 Table 6. Modified Patriarchy Results for Men in the Philippines 48 Table 7. Modified Patriarchy Results for men in the Philippines 49 iv List of Figures Figure 1. Research Model v Acknowledgements To complete a thesis each graduate student must have an exceptional team of individuals who foster and encourage growth. I offer my sincerest thanks to the faculty and staff of Family Studies especially to Dr. S. Marshall who saw a potential academic and planted seeds that changed the course of my life in ways that I am so grateful for. I would also like to offer my deepest gratitude to Mr. D. Wun, Ms. L. Catto, Ms. A. Sokol and Mr. S. Deyell for the acceptance and support they all offered through this process. I have to thank Dr. C. Yodanis for her enthusiasm, wisdom, encouragement, and guidance that she so willingly shared with me. I could not have asked for a better advisor. Finally a special thank you to Jason Gozjolko, my husband and best friend, who saw my potential, who has encouraged my goals, and who continues to believe and support me as I chase my dreams. vi To Jason, without your support none of this would have happened. vii Introduction The majority of marital power literature states that the partner who contributes the most resources to the marriage, such as income, has the greatest amount of marital power (Allen, 1984; Blood & Wolfe, 1960; Brinkerhoff & Lupri, 1989; Gray-Little, 1982; Heer, 1963; Lee & Petersen, 1983; Safilios-Rothchild, 1970; Xu & Lai, 2002; Yodanis, 2003). Most investigations of marital power focus on individual's characteristics within the relationship with little consideration of the affect of the social context in which the relationship exists (Alverez, 1979; Blood & Wolfe, 1960; Conklin, 1979; Elliott & Moskoff, 1983; Xu & Lai, 2002; Zuo, 2004). Rodman (1972) proposed that social context, the dominant gender ideology of a country, does impact marital dyads and that there is more to relationships than simply the individual's characteristics. This investigation will explore how country level gender ideology moderates the relationship between resources and marital power. 1 Literature Review The Importance of Studying Marital Power Marital power, specifically power inequality, is linked to marital dissatisfaction, marital conflict, and intimate partner violence, as well as poor health outcomes, such as higher incidences of depression and greater prevalence of general illness (Felmlee, 1994). Extreme power inequality is also correlated with an increase in marital discord and a corresponding decrease in relationship stability for married couples (Alverez, 1979). The perception of power inequality also impacts individual's reports of marital satisfaction (Gray-Little, 1982). Satisfaction for women in the conjugal relationship is found to be dependent on perceived equality. Yet, male marital satisfaction is associated with perceived power imbalance, or inequality. Husbands who report high power relative to their wives indicate greater marital satisfaction than husbands who report equal or lower power relative to their wives (Gray-Little, 1982). These results are contrary to findings from Amato, Johnson, Booth and Rogers (2003) who report that greater power inequality is correlated with decreased intimacy and lowered marital quality over time. Despite these varied findings, the over all picture of inequality suggests that power impacts the most fundamental processes in the marital relationship such as communication, expression of affection, and family operation (Rodman, 1972). Power Types A general definition of power is necessary to orient the focus of the research. Power for the purpose of this project, is understood in the tradition of Weber (1947) as: power over. This specifically is the "ability to control others, events or resources - to make happen what one wants to happen in spite of obstacles, resistance, or opposition" (Weber, 1947, p. 152). 2 Although a broad power definition is important, problems with previous marital power research have resulted from a lack of definitional specificity (Szinovacz, 1986). Using only a general explanation does not reflect .the dynamic nature of marital power (Brinkerhoff & Lupri, 1989; Scanzoni, 1979; Szinovacz, 1986). Clearly defining the type and dimension of power being measured is important to reporting findings and making appropriate generalizations about marital power. When attempting to create a specific operational definition of marital power, two different types of power emerge from the existing literature: potential power and actual power. Power is often defined in terms of potential (Agarwal, 1997; Brinkerhoff & Lupri, 1989; Cromwell & Olson, 1975b; Rollins & Bahr, 1976; Scanzoni, 1979; Szinovacz, 1986). Potential power is based on the partners' relative perceptions of the ability to influence others (Olson & Cromwell, 1975). Although potential power is important, the concept of actual power is the focus of this investigation. Actual power is defined as the ability of one individual to change another person's behaviour (Cromwell & Olson, 1975b). This concept is operationalized through decision making indicators. Actual power is demonstrated when one participant in the dyad makes a final decision about a situation affecting both members of the dyad (Blood & Wolfe, 1960). This decision is then considered final, regardless of any persuasion that may have been attempted by the non-decision making partner (Szinovacz, 1986). Power Domains Understanding and identifying the dimension, or domain, of power under investigation is essential to developing an appropriate and specific operational definition of power. Originally power was not conceptualized as possessing different domains. Rather, it 3 was assumed that power was a uni-dimensional concept. However as research regarding power has become more sophisticated, power has emerged as a multi-dimensional concept (Centers et A/., 1971; Heer, 1963; McDonald, 1980; Sabatelli & Shehan, 1993; Safilios-Rothchild, 1970; Scanzoni, 1979; Szinovacz, 1986). Three domains of power are identified: power bases, power process and outcome power (Brinkerhoff & Lupri, 1989; Cromwell & Olson, 1975b). Marital power in this investigation is operationalized as outcome power. Outcome power, or the effects of power (Scanzoni, 1979), is the end result, or who affects change in whom. Outcome power is operationally defined as the ability to have the final say, or make the final decision in the relationship. The judgment of who in the relationship is the most powerful is indirectly assessed through individuals' reports of which partner has final say in decision making (Brinkerhoff & Lupri, 1989; Safilios-Rothchild, 1970; Szinovacz, 1986). Final say decision making measures have been the most highly used indicators of marital power (Brinkerhoff & Lupri, 1989). Final say decision making indicators were developed by Blood and Wolfe (1960) with the express purpose of measuring power between husbands and wives in marital relationships. The indicators for the final say decision making scale are based on three criteria (Blood & Wolfe, 1960). The first criterion is that decisions used to assess power should affect all members of the family and should be comprised of issues that all married individuals face. For example, not all conjugal dyads have children, therefore no questions about children are included in the decision making scale. The second criterion for inclusion in the scale is that each decision should be important to family functioning. The final criteria is that masculine and feminine qualities should be excluded from the indicators (Blood & Wolfe, 1960; Centers et al., 1971). The rationale for ensuring 4 masculine and feminine qualities are absent from the questions is to avoid an inherent gender bias. Blood and Wolfe (1960) designed the questions to be gender neutral as not to be interpreted by the respondents as specifically directed towards one gender over another. Blood and Wolfe's (1960, p. 19) original decision making index consists of these eight questions: 1) What j ob should the husband take? 2) What car to get? 3) Whether or not to buy life insurance? 4) Where to go on a vacation? 5) What house or apartment to take? 6) Whether or not the wife should go back to work? 7) What doctor to have when someone is sick? 8) How much money the family can afford to spend on food a week? Criticisms of Final Say Decision Making Indicators These final say decision making measures are strongly critiqued in the literature. Criticism focuses on issues regarding the number of questions used to assess decision making, the types of questions asked, the relative frequency of decisions, and the specificity of decision making that can be measured by the survey research questions (Brinkerhoff & Lupri, 1989; Safilios-Rothchild, 1970; Szinovacz, 1986; Xu & Lai, 2002; Yodanis, 2003). One critique of the final say decision making scale is that all indicators are equally weighted (Safilios-Rothchild, 1970). Simply adding the indicators, to establish a composite marital power score, suggests each indicator is of equal weight. There is no recognition of potential value differences between the indicators. This equal weighting does not take into account the frequency or importance of any one decision (Cromwell & Olson, 1975a; Mizan, 1994; Olson & Rabunsky, 1972). The ability to make one important decision may reveal more about the power structure in a relationship than the ability to make several unimportant 5 decisions and so by equally weighting the decision making indicators an inaccurate reflection of the marital power structure is created (Mizan, 1994; Safilios-Rothchild, 1970). A second criticism of the final say decision making measures scale is that delegation of decisions is not reflected (Safilios-Rothchild, 1970; Szinovacz, 1986). The scale simply captures who makes the decisions. The decision making scale is theoretically based on the assumption that the power rests with the person who makes the decision. This assumption may lead the spouse who is responsible to make the decisions to be incorrectly assessed as the most powerful individual in the conjugal dyad. This incorrect assumption results from the fact that if decisions are seen as boring, time consuming or bearing unwanted responsibility, one spouse may delegate the responsibility of decision making to their partner. Safilios-Rothschild (1970) suggests the person who can delegate decisions is able to orchestrate the marital power structure and therefore possesses the ultimate power. Yet, this ultimate power is not captured by the final say decision making scale. The largest criticism of final say decision making scale is the inability of the indicators to evaluate process orientated interactions that occur before the decision is made (Safilios-Rothchild, 1970; Scanzoni, 1979; Sprey, 1999; Szinovacz, 1986). During the original conceptualization of the decision making scale, Blood and Wolfe (1960) were not interested in investigating the negotiation process of married couples. Process was not of interest because Blood and Wolfe (1960) assumed husbands and wives consulted each other about every decision. Due to this belief in a consultation process, Blood and Wolfe (1960) assumed that the person who made the final decision was the individual who had the power to decide the course of action for the family. Safilios-Rothschild (1970) argues that power process, conceptualized as negotiation, is the more important indicator, and therefore 6 indicators of negotiation should be used to establish who has power in the marital relationship. Negotiation is believed to have greater importance than final say indicators. Decisions are believed to be made during the negotiation process and the person who is indicated to have final say with decision making measures is only the spokesperson for the couple not the most powerful partner (Sprey, 1975). The truly powerful partner is the individual who is able to fulfill their own needs or wants by altering the other person's behaviour choices during the negotiation process. This process of negotiation and influence is not assessed by the final say decision making scale. A final critique of the decision making measures is that the measure is based on the assumption that every person in the marital relationship has the same potential and ability to access power (Komter, 1989). This may not be an accurate assumption, as Komter (1989) suggests that the final say decision making scale only assesses overt power in relationships. Manifest, or overt, power is the only type of power, according to Komter, that can be measured with final say decision making indicators. Latent power, seen as covert power attempts, is not evaluated by decision making measures yet it is an important aspect of marital power (Komter, 1989). Covert power attempts can be considered process orientated because latent power is only observable during negotiation within the marital dyad. The final area of power, as identified by Komter, is hidden or invisible power. This power may involve some of the most interesting dynamics as Komter (1989) suggests that one of the partners, usually the wife, may not even be aware that power is available for her to use. Unlike with latent power, there are no attempts to use influence or conflict to sway or participate in decision making. When hidden power is present in a relationship, Komter (1989) suggests that the wife does not understand or realize there could be an opportunity to 7 access power to express her wants, needs and desires (Komter, 1989). Invisible power is designated to one partner, usually the husband, through social norms and expectations. Thus the assumption that all parties have equal access to marital power is a deficit of the final say decision making measures. Despite the numerous critiques of outcome power and decision making measures, actual power is the most widely researched domain of power, especially in survey research. In most data sets, final say decision making measures are the only indicators available to study marital power. Understanding the critiques of the decision making scale is an important aspect of using the final say indicators to complete research investigations. The importance of continuing marital power research outweighs the inability of the final say decision making measures to access all domains of power. Although it is clear the final say decision making scale needs to be improved, the use of this scale allows marital power research to continue until a new survey measure is developed. Definitions and dimensions of power focus on the individual's characteristics in interactions and do not include the social context within which relationships exist. Relational concepts such as power are affected by social norms, beliefs and values (Alverez, 1979; Bolak, 1995; Conklin, 1979; Connnell, 1987; Felmlee, 1994; Godwin & Scanzoni, 1989; Lavee & Katz, 2002). Authority explains hidden power and demonstrates how social norms and expectations impact the individual's characteristics within social interactions. Authority and Marital Power Authority is an important concept to understand when investigating power in marital relationships (Adams, 2004; Blood & Wolfe, 1960; Brinkerhoff & Lupri, 1989; Burr et al., 1977; Corrales, 1975; Cromwell & Olson, 1975b; Goldhamer & Shils, 1939; Rodman, 1972; 8 Rollins & Bahr, 1976; Scanzoni, 1979; Sprey, 1975). Perelberg (1990) states that "authority is linked to the idea of legitimization, the right to make particular decisions, and to command obedience. Power, in contrast, lies in the possibility of imposing one's will upon the behaviour of other persons. Every society contains multiple sources of both power and authority" (p.44). Authority is the culturally defined legitimate right to possess and assert power and to behave in specified ways, with regards to power, in the marital relationship (Brinkerhoff & Lupri, 1989; Sprey, 1975). Authority conceptually links social context to the individual's power in marital relationships. It is the social norms and values regarding power in marital relationships that designate who should have marital power. One example of authority is the power men access in patriarchal societies. Men in a patriarchal society are designated as the power individuals in the relationship regardless of skill, ability, or contribution to the relationship (Brinkerhoff & Lupri, 1989). Social norms do not designate power to women in patriarchal cultures. Women may not even realize that there is a possibility that power could be structured differently in the marital relationship. These are the norms and social restraints that are identified by Komter (1989) as hidden power. Resource theory To explain marital power, resource theory (Blood & Wolfe, 1960) has been developed. The main hypothesis of resource theory is that the individual who provides the highest number of needed resources to the marital relationship, relative to their partner, will have the most power in the relationship, and this power will be demonstrated by the ability to have the final say in decision making (Allen & Straus, 1984; Blood & Wolfe, 1960; Brinkerhoff & Lupri, 1989; Conklin, 1979; Cromwell & Olson, 1975b; Rodman, 1967; Szinovacz, 1986). 9 Resources are anything, tangible or intangible, which one partner offers, or makes available, to meet the other partner's needs (Blood & Wolfe, 1960; Felmlee, 1994). The person who is able to supply, or withhold, access to resources will be able to exert power within the marriage (Goldhamer & Shils, 1939). Resources can be either internal or external to the relationship. Due to the difficulty in measuring internal resources in survey research, external resources such as relative measures of social status, education, and income have been widely used. The amount of power an individual possesses in the marital relationship is therefore hypothesized to be constrained by the volume and number of valued relative external resources the individual has contributed to the conjugal dyad (Cromwell & Olson, 1975b). Since its conception, resource theory has been used to try to explain marital power in the United States and also cross culturally (Brinkerhoff & Lupri, 1989; Cromwell & Olson, 1975a; Hesse-Biber & Williamson, 1984; Lee & Petersen, 1983; McDonald, 1980; Rollins & Bahr, 1976; Safilios-Rothchild, 1967; Scanzoni, 1979; Szinovacz, 1986; Xu & Lai, 2002). The majority of marital power research has examined power in single nation studies and the findings have supported Blood and Wolfe's hypothesis (Agarwal, 1997; Allen, 1984; Allen & Straus, 1984; Blood & Wolfe, 1960; Centers et al., 1971; Dhruvarajan, 1992; Elliott & Moskoff, 1983; Felmlee, 1994; Gray-Little, 1982; Grinberg-Vinaver, 1955; Hallenbeck, 1966; Heer, 1963; Hesse-Biber & Williamson, 1984; McDonald, 1980; Oropesa, 1997; Scanzoni, 1979). However, when resource theory is applied cross culturally it does not explain marital power in every country context (Rodman, 1967; Safilios-Rothchild, 1967). 10 Country as Context Conklin (1979) reminds readers that resource theory was developed and tested in a single country: the United States. This theory was not specifically created to explain cross national differences in patterns of marital power. Therefore, when investigating marital power cross nationally it is important to consider differences in cultural and social norms, beliefs, and expectations together known as cultural or country context (Conklin, 1979). The country as the context is employed when a researcher is interested in how an aspect of the social structure impacts personality, individual actions and behaviours (Kohn, 1987). Using country as the context allows social norms, beliefs and expectations to be compared between countries to understand how these ideals impact individuals and their interactions. There are some important issues to remember when completing cross national comparative research such as, how the data is collected, who is selected, and the ability of the collected sample to accurately represent the country (Kohn, 1987). When collecting the data, investigators in each country have discretion in the selection of research participants. Research teams may choose to use a convenience sample from a major urban center, while in another country a research team may choose to randomly sample the both urban and rural areas. The choice of sampling procedure influences the comparability of countries within one data set. Longitudinal comparison can also be a challenge when completing country as context cross national research. This type of comparison can be problematic because the individuals in the sample are not necessarily consistent from year to year (Kohn, 1987). A final issue is that often comparative research involves using several different data sets. Researchers attempting to complete cross national research may find that not every country of interest is present in each data set being used. 11 However, the benefits of comparative research are greater than the limitations presented. The ability to compare social context between countries is important for gaining an understanding of how cultural context can impact individuals during interactions. Resources in a Cultural Context Rodman's (1967, 1972) theory of resources in a cultural context builds on resource theory. The hypothesis of resources in a cultural context is that the relationship between resources and power is moderated by the dominant gender ideology of the country in which individuals live. How marital power is structured is impacted by more than simply the interactions between the individual participants in the relationship. Rodman's (1972) hypothesis is that power structure in the marital relationship is not simply decided by relative resource contribution as suggested by resource theory. Resources in a cultural context integrates cross cultural social norms to produce a more complex and comprehensive explanation of the relationship between resources and marital power (Szinovacz, 1986). The theory of resources in a cultural context is built on two specific beliefs; "the distribution of marital power is influenced by the interaction of (1) the comparative resources of husband and wife, (2) the cultural or subcultural expectations about the distribution of marital power" (Rodman, 1967, p. 322). Gender ideology Gender ideology is a set of beliefs and attitudes about gender which individuals hold that have been formed through established cultural and social norms, beliefs about roles, and expectations for men and women (Agarwal, 1997; Alverez, 1979; Brinkerhoff & Lupri, 1989; Elliott & Moskoff, 1983; Felmlee, 1994; Grinberg-Vinaver, 1955; Hallenbeck, 1966). Beliefs, values and expectations about gender are pervasive, ingrained and seem so natural 12 individuals often do not realize the impact gender ideology has on everyday interactions (Greenstein, 1996a). Every exchange within the marital dyad is, therefore, impacted by the dominant beliefs, attitudes and cultural values that individuals hold about gender (Greenstein, 1996b; Nordenmark, 2004; Ross, 1987). It is gender ideology as the cultural context that guides the beliefs about the distribution of power. The dominant gender ideology of a culture specifies which gender should have authority in a society, and also in marriage. Rodman (1972) begins to explain why the contribution of valued resources that leads to greater power within marriage in some countries, does not lead to greater conjugal power in all countries. Rodman suggests that it is the dominant gender ideology as the cultural context that impacts men's and women's definition of who should have power in the marital relationship. He identified three different types of cultural contexts: traditional, transitional and egalitarian gender ideologies (Rodman, 1972). Rodman's Cultural Contexts Egalitarian. A culture with an egalitarian gender ideology stresses sharing power between partners and has strong family norms regarding equality between husbands and wives (Rodman, 1972). These egalitarian norms have been so well established that resources do not affect marital power (Rodman, 1972). In an egalitarian culture, no relationship is hypothesized to be found between resources and power as a result of the strong gender norms. It is expected that regardless of resource contribution, each member of the conjugal dyad will have equal access to marital power. 13 Traditional. In a traditional culture, there are strong patriarchal norms and a high level of male authority. Men in a traditional country are granted authority and hold the power positions in society. Men are the leaders of the family regardless of their status as income earners (Rodman, 1967, 1972; Warner et al., 1986). If a country is strongly patriarchal, it does not matter how many resources a woman brings to the marital relationship. Even with high resource contribution, relative to her husband, the wife will not be able to access marital power. This inability to access power is due to the strong cultural norms which prevent male authority from being challenged (Rodman, 1972). Regardless of the volume or number of relative resources these men contribute, the cultural norms of paternal authority keep them in a power position within the family and the larger culture (Rodman, 1972). In traditional cultures, Rodman (1972) discusses a second type of gender ideology identified as "modified patriarchy". Modified patriarchy is found in countries where the lower classes still adhere to the traditional patriarchal gender ideology but the upper classes have an egalitarian gender ideology. In these countries male authority is inversely related to social class, but it is not that the men in these countries have less patriarchal authority (Rodman, 1972). What is different is that the men in the upper classes are shifting and adopting egalitarian norms therefore modifying the "patriarchal tradition"(Rodman, 1972, p. 64) Transitional. A transitional culture is "a society in which equalitarian family norms are replacing patriarchal norms in which there is normative flexibility about marital power" (Rodman, 1972, p. 64). In these countries there is more confusion and questioning of social norms and cultural values as these beliefs and ideals are in the process of changing. In 14 transitional societies, Rodman (1972) suggests that relative resource contribution will influence the acquisition of marital power. He hypothesizes that the contribution of valued resources will mean more access to marital power, and this access will mean the ability to make decisions. Rodman (1972) also suggests that in transitional countries, power and esteem are not assigned based on position in society, but are achieved by attaining and contributing resources. Contributing resources, such as higher relative educational status and greater relative income, to the marital relationship allows the contributor to access more marital power. In transitional cultures marital power is not assigned through norms, beliefs, and values. If men are not the primary breadwinners it is possible that the ultimate power may be possessed by the woman (Rodman, 1972). The majority of empirical tests of resources in a cultural context have been single nation research investigations (Agarwal, 1997; Blood & Wolfe, 1960; Bolak, 1995; Centers et al., 1971; Gray-Little, 1982; Greenstein, 1996a, 1996b; Hesse-Biber & Williamson, 1984; Lavee & Katz, 2002; Min, 1997; O'Connor, 1991; Webster, 2000; Xu & Lai, 2002; Zuo, 2004). In single nation studies rather than investigating gender ideology as the cultural context, an individual's gender ideology is tested. This has led to myriad of findings which have been inaccurately used to support or refute Rodman's theory of resources in a cultural context. Using multiple countries and comparative data, dominant gender ideology can be studied as the context for the marital relationship and therefore the relationship between resources and power can be compared across contexts to test with more accuracy Rodman's theory of resources in a cultural context. 15 Hypotheses This study will attempt to discover if the power structure in marriage is affected by the external relative resources that each individual in the conjugal dyad contributes to the marital relationship. A second objective of this investigation is to discover if the relationship between resources and marital power is moderated by country context. This study will attempt to address the following hypothesis: HI a: Men will have the most power in traditional cultures. Hlb: Men and women will have the most equal power in egalitarian cultures. H2a: As women's relative resource contribution increases, women's marital power increases. H2b: As men's relative resource contribution increases, men's power increases. H3a: The cultural context within which a relationship exists moderates the relationship between relative resource contribution and marital power. H3b: In the egalitarian context, as men's and women's resource contribution increases, no corresponding increase or decrease is observed in marital power. H3c: In the traditional context, as men's and women's resource contribution increases, no corresponding increase or decrease is observed in marital power. H3d: In the transitional context, as men's and women's resource contribution increases, marital power increases. H4: In the traditional context, as family income increases, there will be a corresponding increase in gender ideology (modified patriarchy). 16 Figure L Research model. Gender Ideology Relative income Relative education 17 Methods To test the hypotheses, the 2002 International Social Survey Program (ISSP) Family and Changing Gender Roles III, a cross national data set, is used in this study. The ISSP was created to permit empirical research with equivalent data responses from a selection of countries (Diefenbach, 2002). Data sets such as the ISSP increase the ability to compare across countries with standardized indicators through self report survey, thereby allowing greater accuracy in comparative research. The 2002 ISSP family and changing gender roles III is composed of 35 different countries. The countries included are Austria, Australia, Brazil, Bulgaria, Chile, Cyprus, Czech Republic, Denmark, Finland, Flanders (Belgium), France, Germany, Great Britain, Hungary, Ireland, Israel, Japan, Latvia, Mexico, Netherlands, New Zealand, Northern Ireland, Norway, Philippines, Poland, Portugal, Russia, Slovenia, Slovakian Republic, Spain, Sweden, Switzerland, Taiwan, and the United States. There are 26,847 married or living as married respondents in this data set. Country Context Cross national research is any investigation that compares between two or more countries using equivalent data (Kohn, 1987). This investigation is designed to explore how gender ideology moderates the relationship between relative resources and marital power across countries. If the relationship between relative resources and marital power is different in two countries, it is possible that gender ideology as the social context is a moderating variable. Cross national comparison is not an area that has been greatly investigated in marital power research. Constructing the gender ideology scale is an important step toward establishing each country's context as traditional, transitional or egalitarian. Previous marital power research 18 has simply assigned gender ideology to a country based on assumptions about the culture rather than creating a scale to objectively categorize gender ideology. In this study a gender ideology scale is created to attempt to establish an accurate aggregate gender ideology score for each country. Previous research has employed one to four variables to establish a mean gender ideology score (Diefenbach, 2002; Greenstein, 1996a; Nordenmark, 2004). Diefenbach (2002) used one indicator to establish gender ideology. However, one indicator seems to be questionable in establishing an accurate aggregate gender ideology score. In the present investigation four indicators are used to create the gender ideology scale (Nordenmark, 2004). The four variables that are used to create the composite gender ideology scale are: 1) All in all family life suffers when the woman has a full time job; 2) A job is alright, but what most women really want is a home and children; 3) Being a housewife is just as fulfilling as pay; 4) A man's job is to earn money and a woman's job is to look after the home and family. This scale is a composite scale as the indicators are summed to create a total gender ideology score. The answer choices are, strongly agree, agree, neither agree nor disagree, disagree, strongly disagree. To minimize missing respondents, an average has been taken of all the responses for each participant. The minimum number of responses necessary to be included in the analysis is three. Individuals who have two or less responses are considered missing and are not included in the analysis. Respondents receive a score from the scale of 1-5. A low score indicates a patriarchal response and a high score indicates an egalitarian response. 19 Three countries from the 2002 ISSP data set have been chosen to complete this investigation. Each cultural context, traditional, transitional and egalitarian, is represented by one country. Within the framework of this investigation it is important to ensure that the country chosen to represent the transitional context can actually be considered to be in transition. To ensure an appropriately chosen transitional country, previous ISSP Family and Changing Gender Role data sets were accessed. Two data sets available for use to assess if countries are transitional are the 1988 ISSP Family and Changing Gender Roles I and the 1994 ISSP Family and Changing Gender Roles II. Both the 1988 and the 1994 ISSP Family and Changing Gender Roles data sets have all of the necessary variables available to create the gender ideology scale. The participant countries available in the 2002 ISSP are not all present in the 1988 and the 1994 data sets. However, all of the countries that are possible candidates to represent the transitional context are present in all three ISSP Family and Changing Gender Roles data sets (See Appendix 1). Due to the lower number of participant countries in the 1988 and the 1994 ISSP these earlier data sets are only used to assess change in the aggregate mean gender ideology in countries being considered to represent the transitional context. Assessing the aggregate mean gender ideology of the candidates for the transitional country at two earlier points in time assists in establishing change or stability in the mean aggregate gender ideology score between 1988 and 2002. The country that is chosen to represent the transitional context is the United States. The transitional country is a country that demonstrates a progression from 1 to 5 on the gender ideology scale between 1988 and 2002. Ireland is immediately a clear choice for the transitional country as it has a linear progression from a score of 2.76 in 1988 to a score of 20 2.90 in 1994 and finally to a score of 3.15 in 2002. Despite this almost model progression Ireland is ineligible for the analysis because Ireland is missing the key variable of education level for the spouse. The United States is the next most theoretically sound choice. There are qualitative and quantitative reasons for choosing the United States as the transitional country. The United States has also been the focus of much of the research on marital power (Brinkerhoff & Lupri, 1989; Scanzoni, 1979; Szinovacz, 1986). Researchers have often contextualized the United States to represent egalitarian gender ideology (Greenstein, 1996b). The results from the gender ideology scale in this investigation indicate that using the United States to represent the egalitarian context may not accurately reflect the cultural context. A quantitative reason for using the United States as the transitional country is that the American data has all of the necessary response indicators for the analysis. Although the United States appears to have a smaller transition from 3.06 in 1988 to 3.11 in 1994 and finally to 3.09 in 2002 it still scores as a transitional country. Understanding the data from a qualitative perspective allows this movement towards egalitarian and backwards toward traditional gender ideology in the United States to be understood as demonstrating the flexible and uncertain gender norms captured in a transitional country context. The countries chosen to represent the traditional and egalitarian context are present in both the 1994 and 2002 ISSP to ensure the countries are stable non-transitioning countries. Stability in mean gender ideology is represented by a constant aggregate score on the gender ideology scale between the 1994 and 2002 ISSP. The country chosen to represent the egalitarian context is Sweden. Sweden and Norway are the highest scoring countries present in both 1994 and the 2002 ISSP data sets. Sweden is chosen because it appears to have the most stable aggregate gender ideology with 21 a score in 1994 of 3.34 and a score of 3.39 in 2002. The reported mean gender ideology score for Norway changed by 0.19 points, lacking the stability of the aggregate gender ideology score reported by Sweden. The country chosen to represent the traditional context is the Philippines. In 1994, aggregate mean gender ideology score for the Philippines is 2.39 and in 2002 the score is 2.40. The Philippines is chosen due to the stability of the aggregate gender ideology score between the 1994 and 2002 ISSP data sets. See Table 1 for sampling and data collection procedures. Table 1 Sampling and Data Collection Procedures Country Year Gathered Sampling Sample Size Collection Sweden 2002 Probability 1171 In person self administered survey USA 2002 Probability 1080 Mailed survey Philippines 2002 Probability 1200 Face to face interviews Dependent Variable Decision making scale. The ISSP only has two final say decision making variables available to measure marital power. The individual questions are: When you and your spouse/partner make decisions about the following, who has the final say when: 1) Choosing shared weekend activities; 2) Buying major things for the home. The response categories are: mostly me, mostly my spouse/partner, sometimes me/sometimes my spouse/partner, we make decisions together, someone else. The response of someone else is not used in this analysis as it does not address outcome power in the marital dyad. Not using the response someone else results in approximately 100 cases being excluded from the entire sample. 22 The original response categories do not reflect the respondent's gender therefore the categories are recoded to reflect gender. Next, a composite power variable is created using these two decision making indicators. Factor analysis indicates that these two variables hang together in the same component with a factor score of .80 and a Cronbach's alpha of .67 so the creation of a scale variable is appropriate. The categories of sometimes man/sometimes woman decides and decides together have been recoded to form one category as both the individual categories indicate shared decision making. This results in the recoded composite power variable having three categories. The new gender obvious categories are men's decision; women's decision and shared decision. This indirect measure of power demonstrates a non-normal distribution. The non-normal distribution could not be transformed due to the high percentage of respondents who reported shared decision making in the marital relationship. Table 2 shows that 66.3% of the respondents indicate shared decision making. Combined only 33.7 % of the participants indicate either of the remaining two categories of men's decision or women's decision. This trend is seen occurring in each country and it is still present when the sample is divided by gender into men and women. The non-normal distribution results in the need for a nominal variable to enable multinomial logistic regression analysis. The categories of interest in the recoded power variable are men's decision and women's decision and the reference category for the variable is shared decision. Independent Variables Resources. Resources are the independent variables for the analysis. There is some controversy about which variables should be used to measure resources in marital dyads (Allen, 1984; Allen & Straus, 1984; Bahr, 1972; Blood & Heer, 1963; Blood & Wolfe, 1960; 23 Burr et al, 1977; Centers et al, 1971; Diefenbach, 2002; Elliott & Moskoff, 1983; Hesse-Biber & Williamson, 1984; Lee & Petersen, 1983; Mizan, 1994; Olson & Rabunsky, 1972; Scanzoni, 1979; Sprey, 1972; Szinovacz, 1986; Xu & Lai, 2002). Researchers suggest that only external resources, such as income and education, should be used when investigating the relationship between resources and marital power (Adams, 2004; Allen & Straus, 1984; Blood & Heer, 1963; Blood & Wolfe, 1960). Alternatively, a combination of internal and external resources are proposed as a better indicator of resources (Blood & Heer, 1963; Heer, 1963; Safilios-Rothchild, 1970). It is also believed that the best measure of resources is not a measure that uses the absolute amount of resources each partner acquires, but rather an indicator that uses the resources that an individual contributes relative to their dyadic partner (Blood & Heer, 1963; Burr et al., 1977; Centers et al., 1971; Heer, 1963). In this research, external relative resources are measured. The variables which are used in this study are relative income, and relative education. Relative income. The indicator question is "Considering all of your sources of income between you and your partner who has the higher income?" This variable ranges from 1-7 indicating different relative income arrangements between the marital partners. This variable has response categories of: man has no income, man has much higher income, man has higher income, equal or about the same income, woman has higher income, woman has much higher income, woman has no income. This variable is recoded to reflect which partner, the man or the woman; the respondent indicates is the main income contributor relative to their spouse. The recoded response categories are: woman has no income, woman has much lower income, woman has lower income, woman has equal or about the same income, woman has higher income, woman has much higher income, woman has only 24 income. A low score on this variable indicates the woman contributes less income relative to her partner or that the man has a higher relative income. A high score indicates the woman contributes more income relative to her partner or that the man has a lower relative income. Relative education. Education is another resource that influences power in the marital dyad (Adams, 2004; Bahr, 1972; Blood & Heer, 1963; Blood & Wolfe, 1960; Burr et al., 1977; Centers et al., 1971; Cromwell & Olson, 1975b; Dhruvarajan, 1992). Bahr (1972) suggests that it is not the man's or woman's education level that predicts ability have the final say or the outcome power in the marital relationship but the education level one partner has in relation to the other partner in the dyad. In previous research a positive relationship has been found between the husband's education and the husband's power (Bahr, 1972). Earlier studies have found that it is not the husband's education which matters in isolation but the relative education between the husband and the wife (Bahr, 1972). The education variable, highest education level, will be used as it is the only education indicator that has both respondent and partner information available. The response categories are: no formal education, lowest formal education, above lowest formal education, higher secondary education completed, above higher secondary education completed below university, university degree completed. The education variable is recoded to create a relative education variable. The spouse education level is subtracted from the respondent education level creating a positive or negative number. This information is then recoded to create two dummy variables respondent has higher education level than spouse and respondent has lower education level than spouse. Equal education level is the reference category for both higher education level and lower education level. 25 When men and women are analyzed together in the regression analysis, the education variable measured differently than described above. When men and women are analyzed together only the higher education dummy variable is used to create a relative higher education level variable. The dummy variable is coded to understand if the respondent is indicated that the man or the woman has a higher education level. After this recoding is complete then the variable is recoded into a three category nominal variable of man has higher education level, woman has higher education level and equal education level. Control Variables Gender ideology. Gender ideology is an important aspect of this research project. Gender ideology is used as a control variable to ensure that when measuring the relationship between resources and marital power each individual respondent is considered to have the aggregate mean gender ideology score. This variable is a composite measure. The four variables included in the composite measure are: 1) All in all family life suffers when the woman has a full time job; 2) A job is alright, but what most women really want is a home and children; 3) Being a housewife is just as fulfilling as pay; 4) A man's job is to earn money and a woman's job is to look after the home and family. The answer choices are, strongly agree, agree, neither agree nor disagree, disagree, strongly disagree. The variable score ranges from 1 to 5. A low score indicates a traditional response and a high score indicates an egalitarian response. To minimize missing respondents an average is taken of all the responses for each participant. The minimum number of responses necessary to be included is three. Individuals who had two or less responses are considered missing and are not included in the analysis. 26 Age. It is important to control for age in this research project because age is found to impact the relationship between resources and power in marital dyads. As age increases the power structure seems to become more equal as it is found that older couples report more egalitarian power in their marriage (Hesse-Biber & Williamson, 1984). It is possible that this may occur because of the stage of the life cycle these older couples have reached. Previous research has found that older couples are more likely to report sharing housework and also sharing decision making (Hesse-Biber & Williamson, 1984; Szinovacz, 1986). This suggests marital power may be more egalitarian later in the life cycle. Age is a scale variable as each respondent entered the appropriate age rather than choosing an age category. Children in the house. Having children in the household impacts marital power (Blood & Wolfe, 1960). The impact could be due to more traditional roles which have been found to develop when children enter the family environment (Greenstein, 1996a). The arrival of children often leads to one parent remaining at home with the child for a period of time. It is often the woman who remains at home with the child. The wife's lack of access to external resources may lead to a diminished relative resource contribution that may explain the shift to traditional roles that has been found (Blood & Heer, 1963; Blood & Wolfe, 1960). By controlling for the presence of children in the home it is possible to look at the affects that the presence of a child has on marital power. To control for the presence of children, the household composition variable is recoded. This nominal variable asks respondents to indicate how many adults and children live in their primary residence at the time of the survey. The household composition variable gave options for respondents from 1 to 12 adults living in the house and 1 to 12 adults plus children living in the house. The answer choices are recoded that so any response that 27 includes children is coded as 1 or children present in the household. The reference category 0 is no children present in the household. Marital Status. Only the married and living as married respondents are included in this investigation. There is no variable that isolates the respondents who may be living common law from those who are married. The dataset has been filtered to only include the respondents who are married or living as married and to ensure that no other respondents are added to the analysis. Gender. Gender is an important aspect of marital power research. It is important because many investigations in the literature have found a reporting difference for men and women about men's and women's power in marriage (Agarwal, 1997; Alverez, 1979; Blood & Wolfe, 1960; Bolak, 1995; Conklin, 1979; Corrales, 1975; Sabatelli & Shehan, 1993; Sprecher & Felmlee, 1997; Zipp et al, 2004; Zuo, 2004). When male and female participants are measured together in the analysis a dummy variable of gender is placed in the regression. Women are coded as 1 and men are coded as 0. Men and women will also be analyzed separately to enable a clearer understanding of possible gender differences in reporting. Family income. Family income is an important aspect of power research. Couples in lower income brackets have, indicated in past research, more patriarchal or traditional, patterns of interaction (Rodman, 1972). Higher levels of traditionalism, for lower income couples, may impact who makes the final decision about family purchases and weekend activities (Szinovacz, 1986). Family income is a problematic variable when completing cross national research because income can not be compared as a 1:1 ratio between countries. The inability to compare directly across countries is due to currency differentials. 28 To be able to compare across countries the variable family income is recoded into deciles. Because of differences in wealth between countries this measure is expected to create an ability to compare between the countries. Family income in each country is divided into ten categories and each family in each country is placed in the appropriate category. This means that top, middle and bottom income earners for each country are classed into an appropriate decile that reflects their earning for their country. This is expected to account for any currency differences between countries. The individuals who are in the top income earner bracket for each country are classed into the top percentile category. The 1st decile is composed of the families in each country with the lowest family income and the 10th decile is composed of the families in each country with the highest family income. Modified Patriarchy When testing the modified patriarchy hypothesis a final variable is added to the analysis. This variable is an interaction variable. The interaction is between family income and relative income. To create this variable first family income is recoded into a dummy variable where the respondents who indicated an income above median are the reference category a 0 or the high family income group, and the participants who indicated an income below the median are recoded as the category of interest a 1 or the low family income group. Five is the median for this variable. Relative income is coded in the same way as described above when combined with the recoded dummy family income variable. 29 Analysis The analysis is completed using multinomial logistic regression. This method of analysis utilizes a three category nominal dependent variable with two categories of interest and one reference category. This allows both of the categories of interest in the marital power variable, men's decision and women's decision, to be compared against shared decision. When this nominal power variable is placed in the multinomial logistic regression analysis essentially two logistic regressions are completed. One regression consists of the independent variables men's decision versus shared decision and the second regression consists of the independent variables on women's decision versus shared decision. Multinomial logistic regression is not a linear relationship model such as OLS regression. Multinomial logistic regression uses an odds ratio to express the odds of one event happening over another event. The odds ratio demonstrates how the presence an independent or control variable in the model increases or decreases the odds of the dependent variable occurring. In this model, the odds ratio expresses the odds of men's decision occurring over shared decision and the odds of women's decision occurring over shared decision. The Philippines is the only country in this investigation in which the hypothesis of modified patriarchy is tested using multinomial logistic regression. Two models are completed to test for modified patriarchy in this analysis. The first regression model tests men and women together without the interaction variable. The second regression model includes the interaction variable in analysis to test the impact that family income has on the relationship between resources and marital power. 30 Results Sample Description Control Variables As seen in Table 2 there are 1964 total participants. The Philippines has the highest number of respondents followed by the United States and finally Sweden. There are more female than male participants in each country. The average age for all participants is 46 years old. In general, the men are older than the women in the sample with the oldest average age being found in Sweden and the youngest average age found in the Philippines. A greater percentage of respondents report the presence of children rather than having no children in the household. When this is broken down by country, it is seen in Table 2 that the Filipino respondents report the presence of children more often than the respondents from the United States or Sweden. It is also interesting that fewer participants in the United States and Sweden report the presence of children and a greater number of participants report having no children in the household. 31 Table 2 Descriptive Statistics and Control Variables and Analysis of Mean Differences All Countries Sweden USA Philippines Overall Women Men Overall Women Men Overall Women Men Overall Women Men Respondents 1964 1072 892 532 291 241 560 325 235 872 465 416 Dependent Variable Power % man makes the 14.4 13.8 15.0 5.8 6.3 5.1 19.5 17.3 22.5 16.4 16.1 16.7 decision woman makes the 19.3 22.4 15.6 13.6 15.1 11.8 14.5 15.4 13.2 25.9 31.8 19.2 decision shared decision 66.3 63.8 69.4 80.6 78.5 83.1 66.0 67.3 64.3 57.7 52.0 64.1 making l2 98.68* 67.26* 40.02* 18.41* Independent Variables Relative Education % higher education 25.6 25.7 24.9 28.2 30.0 26.0 25.9 24.3 28.1 23.9 25.0 22.6 lower education r2 25.8 26.2 25.4 26.6 27.1 26.0 26.6 29.3 22.9 24.9 23.5 26.4 Relative income Mean 2.88* 2.89* 2.88* 3.37 3.49* 3.22 3.15 3.39* 2.82 2.39 2.12* 2.69 SD 1.65 1.68 1.62 1.25 1.39 1.03 1.52 1.61 1.33 1.83 1.61 2.02 Control Variables Kids % has kids 59.5 59.9 58.4 40.7 40.6 40.9 38.6 40.7 35.6 84.1 86.8 81.0 no kids 40.5 39.1 41.6 59.3 59.4 59.1 61.4 59.3 64.4 15.9 13.2 19.0 Age Mean 45.92 44.16 48.03 53.24 52.1 54.61 46.30 45.17 47.85 41.21 38.37 44.33 SD 14.40 14.38 14.15 13.43 13.26 13.52 14.73 14.79 14.55 12.78 12.01 12.88 Family Income Mean 6.05 6.10 5.98 6.41 6.57 6.24 6.79 6.98 6.52 5.36 5.22 5.52 SD 2.77 2.74 2.80 2.59 2.64 2.50 2.49 2.37 2.63 2.87 2.78 2.90 Gender Ideology Mean 2.86* 2.92* 2.78* 3.39 3.45* 3.30 3.09 3.23* 2.90 2.40 2.37 2.44 SD 0.90 0.94 0.86 0.86 0.85 0.86 0.94 0.96 0.86 0.65 0.63 0.67 * p = < 0.05 Table 2 shows that when investigating respondents' individual gender ideology the most egalitarian mean score is reported by the Swedish respondents and the most traditional score is reported by the Filipino respondents. There is a significant difference in mean gender ideology scores between countries for all the participants together (F = 272.80, df= 2). The results indicate a significant difference in mean gender ideology scores between countries for men (F = 92.84, df = 2) and for women (F = 189.30, df = 2). The United States is the only country that has a significant difference in gender ideology between men and women (F = 17.16, df = 1). As previously noted it is important to remember that this investigation is not a dyadic measure of husbands and wives, but rather male and female respondents reporting for himself or herself and their spouse. Independent Variables Relative education. Over all 50% of the respondents in the sample report having the same education level as their spouse. When education level is not the same respondents are more likely to report that their partner has a higher level of education. However, it is found in the Philippines that a greater percentage of Filipino women report having a higher education level than their spouse. This trend is also seen in American men where a greater percentage of male participants report that they have a higher education level than their partner. This interestingly does coincide with the reports from the American female participants despite the fact that the respondents are individually measured. A higher percentage of American women report a lower education level than their spouse. Yet, there are no statistically significant differences found for education level between men and women over all, between men and women within each country, or between men across countries, women across countries, or between countries. 33 Relative income. As seen in Table 2, the majority of the sample report that women contribute less relative income to the marital dyad. It is seen that there is a significant difference in the mean of relative income between the countries (F = 69.80, df = 2). Slightly less than 30% of the sample report women contribute equal or more relative income to the marital relationship. There is no significant difference over all between men and women in the mean score for relative income. However, there is a significant difference across countries for women (f = 89.07, df = 2) and for men (F = 7.949, df = 2). It is found that Swedish women have an average relative income that is closest to being an equal income contribution. Only 1.5 % of all Swedish participants report that the woman has no income contribution. However, 54.8% of Filipino respondents report that the woman contributes no income to the marital dyad. It is found that 60 % of female Filipino respondents report having no income. When comparing genders within each country it is seen that the United States (F18.08, df = 1), Sweden (F = 5.95, df = 1) and the Philippines (F = 19.62, df = 1) all have a significant difference in mean relative income scores. Dependent Variable Power. It is seen in Table 2 that 66.5% of all the respondents report that decision making is shared in the marital relationship. Of the respondents that do not report that decision making is shared it is reported more often that women make the decision in the conjugal dyad. All together the Filipino respondents report that women make the decision most often after shared decision making. Table 2 shows that 31.8% of the Filipino women report making the decision in their marital relationship. This trend is reversed in the United States where both men and women report that the man makes the decision more often than the woman but less often than shared decision making. Overall a significant difference is 34 found between countries (%2 = 98.67, df = 2) and between men and women (j2 = 13.64, df= 2). However the findings become clearer regarding power when looking within countries and between the genders. The only significant difference is found in the Philippines where there is a greater likelihood that a female respondent will report that women make the decision in the marital relationship (x2 =18.41, df = 2). Logistic Regression Analysis All Participants The first set of findings to be reported are the results for all of the participants together. When looking at men and women together it is important to remember that this study is not a dyadic level analysis. Each respondent in this study reports for self and their perception of the correct answer for their spouse. Since it is an individual level analysis, two individuals from one marriage are not measured in this investigation. Looking at the male and female respondents together and separately creates a picture of how individuals in marriage perceive power and power relationships but not how men and women in the same marriage may perceive the same relationship differently. All countries. When all the countries are analyzed together several significant findings emerge from the data as seen in Table 3. In regards to men's power what is found is that all the variables that impact men's decision making decrease the odds men will have power in the marital relationship. Relative income is the only relative resource to have a significant affect on men's power. According to men and women, the more income women contribute to the relationship the lower the odds the men will have marital power. For each one unit increase in relative income there is a decrease in the odds that the man will make the decision by 0.918 or 8%. Table 3 shows the finding that gender ideology decreases the odds 35 that men will make the decision in the marital relationship by 0.811 or 19%. This suggests ai gender ideology increases from patriarchal to egalitarian, men's power in the marital relationship decreases. Age also decreases the odds that the man will have power in the marital relationship. For each one year increase in age there is a corresponding decrease in the odds that men make the decision by 0.978 or 2%. The final variable to have a significant impact on men's power is the presence of children. The presence of children in the house decreases the odds that men will make the decision in the marital relationship by 0.741 or 26%. 36 Table 3 Male and Female Participants Multinomial Logistic Regression All Countries Sweden USA Philippines Men's power b Exp(B) b Exp(B) b Exp(B) b Exp(B) Intercept 0.434 0.221 1.147 -1.439 Gender ideology -0.210* 0.811 -0.302 0.739 -0.187 0.829 0.095 1.099 Family income 0.019 1.019 0.087 1.091 -0.028 0.972 0.021 1.021 Relative income -0.085* 0.918 -0.050 0.951 -0.076 0.927 -0.088 0.916 Age -0.022*" 0.978 -0.045* 0.956 -0.021* 0.979 -0.001 0.999 Has kids -0.299* 0.741 -0.727 0.484 -0.493* 0.611 0.042 1.043 Gender -0.013 0.987 0.478 1.613 -0.135 0.874 0.047 1.048 Man higher 0.046 1.047 0.873 2.393 -0.331 0.718 0.125 1.133 education Woman higher -0.299 0.741 0.255 1.290 -1.112* 0.329 -0.012 0.988 education Equal education 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Women's power Intercept -0.738 -3.219 -1.560 -1.158 Gender ideology -0.212" 0.809 -0.007 0.993 0.101 1.106 -0.132 0.877 Family income -0.039 0.962 0.007 1.007 -0.113* 0.893 0.000 1.000 Relative income -0.050 0.951 0.030 1.031 0.062 1.064 -0.049 0.952 Age -0.003 0.997 0.012 1.012 0.003 1.003 0.007 1.007 Has kids 0.354* 1.424 0.524 1.689 0.247 1.281 -0.027 0.974 Gender 0.553*** 1.739 0.470 1.600 0.010 1.010 0.874 *" 2.396 Man higher 0.381* 1.464 1.061 2.889 -0.180 0.835 0.394 1.483 education Woman higher 0.075 1.078 0.100* 1.105 0.199 1.220 0.035 1.036 education Equal education 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 *p = <0.05 **p = <0.01 ***p=< 0.001 #p = <0.1 —I Over all women's power is impacted differently in the model than men's power as seen in Table 3. Direct contribution of resources does not affect women acquiring marital power. Education level is the only variable seen to affect women's power. What is found is that when all countries are analyzed together, when men have a higher education level there is an increase in women's power. Men having a higher education level increases the odds that women will make the decision by 1.464 or 46%. Gender ideology decreases the odds that women will have power in the marital relationship. As gender ideology becomes more egalitarian the odds decrease by 0.809 or 19% that women will make the final decision. The presence of children in the household increases the odds that women will have power in the marital relationship. Having children increases the odds that women will make the decision by 1.424 or 42%. As seen in Table 3, there is a significant difference between men and women when the respondents are reporting about the woman having marital power. When female participants report about marital power the odds increase by 1.739 or 74% that the woman will be reported as the decision maker. The significant difference in how women's power is reported by men and women suggests men and women should be analyzed separately to be able to understand reporting differences between men and women. Sweden. In the egalitarian context it is expected that relative resources will not increase or decrease the odds that men or women will have power in the marital relationship. No relative resources are found to impact man's power in the conjugal relationship. However the respondent's relative education level is found to impact women's power in the marital dyad. Women's power is seen to be increased when the husband has a higher education level than the spouse. A significant difference in education level increases the odds that women will make the decision by 2.889 or 189%. As seen in Table 3, it is found 38 that age affects men's power in the conjugal relationship in Sweden. As age of the respondents increase it is seen that men's power in the marital relationship significantly decreases. For each one year increase in age there is a corresponding decrease in the odds that men will make the decision by 0.956 or 4%. Philippines. In the traditional context it is expected that relative resources will not increase or decrease the odds that men or women will have power in the marital relationship. It is found that relative resources do not have a significant relationship with men's or women's marital power. However there is a significant difference between men and women when reporting women's power. The findings in Table 3 indicate that female respondents have greater odds of reporting that women have the marital power. Female respondents have increased odds of 2.396 or 140% of reporting that women make the final decision. This finding reinforces that it is important to measure men and women separately in order to gain greater understanding of the differences between men and women's reporting of power in the marital relationship. United States. In the transitional context it is expected that relative resources will increase the odds that men and women will have power in the marital relationship. Yet, what is found for men is that each of the variables that impact men's power decrease the odds that men will have power in the marital relationship (seen in Table 3). Relative education is found to have a significant impact on men's power. It is found that when women have a greater education level in the marital relationship there is a decrease in the odds by 0.329 or 67% that men will make the decision. Age is also seen to significantly decrease men's power. As seen in Table 3, each one year increase in age decreases the odds that men will make the decision by 0.979 or 2%. The presence of children also significantly decreases the 39 odds that men will have power in the marital relationship. As seen in Table 3, children in the relationship decrease the odds that men make the decision by 0.611 or 39%. The findings do not indicate that relative resources significantly increases or decreases the odds that women will have power in the marital relationship. However, the findings do indicate for each decile increase of family income there is a corresponding decrease in the odds by 0.893 or 11% that women will make the final decision. Male Participants First, the analysis for men in all of the countries together will be reported. This will be followed by the findings for men in each country individually. It is expected that this analysis will find that there is no significant relationships between relative resources and marital power in Sweden and in the Philippines. It is also expected that there will be a significant relationship between relative resources and marital power in the United States. All countries. Table 4 shows that when men from all countries are measured together the only resource that has a significant relationship with marital power is relative education level. It is found, when men have a higher education level it significantly increases the odds that women will have the marital power. A difference in higher education level significantly increases the odds that women will make the decision by 1.610 or 61%. Two other significant relationships are found. For men, age is found to significantly decrease the odds that men will have marital power. The findings reveal that each one year increase in age decreases the odds that men will make the final decision by 0.978 or 2%. Gender ideology is also found to have a significant affect on the odds of women having marital power for men. As men's gender ideology becomes more egalitarian there is a decrease in the odds that women will make the decision in the marital relationship by 0.795 or 21%. 40 Table 4 Male participants Multinomial Logistic Regression All Countries Sweden USA Philippines Men's power b Exp(B) b Exp(B) b Exp(B) b Exp(B) Intercept 0.184 -1.454 2.223* -1.933 Lower Education 0.081 1.084 0.174 1.190 -0.173 0.841 0.252 1.286 Higher Education 0.087 1.091 1.325 3.763 -0.423 0.655 0.210 1.233 Relative Income -0.071 0.932 -0.026 0.974 -0.273* 0.761 -0.004 0.996 Gender Ideology -0.196 0.822 0.282 1.326 -0.188 0.829 -0.093 0.911 Has Kids -0.239 0.788 -0.602 0.548 -0.769* 0.464 0.302 1.352 Family Income 0.037 1.037 0.263 1.301 -0.051 0.950 0.014 1.014 Age -0.023" 0.978 -0.088" 0.915 -0.026* 0.974 0.010 1.010 Women's power Intercept -1.188* -4.656" -0.615 -0.802 Lower Education 0.257 1.293 0.744 2.104 0.232 1.261 -0.023 0.977 Higher Education 0.476* 1.610 1.289" 3.629 -0.110 0.896 0.416 1.516 Relative Income -0.018 0.983 -0.168 0.845 0.010 1.010 0.014 1.014 Gender Ideology -0.229* 0.795 0.029 1.030 -0.274 0.760 -0.134 0.875 Has Kids 0.193 1.213 0.511 1.668 0.357 1.430 -0.255 0.775 Family Income -0.011 0.989 0.114 1.121 -0.152* 0.859 0.007 1.007 Age 0.003 1.003 0.030 1.030 0.012 1.012 -0.002 0.998 *p = <0.05 **p = <0.01 ***p = <0.001 #p = <0.1 Sweden. In the egalitarian context it is expected that relative resources will not increase or decrease the odds that men will have power in the marital relationship. For men in Sweden it was found that there are no significant relationships between relative resources and marital power that increase or decrease men's power. Although, as it is seen in Table 4, significant relationship between the odds that men will have power and age. It is seen in this context that as men age there is a significant decrease in the odds that the man will have marital power. The odds ratio demonstrates that as age increases there is a decrease in the odds by 0.915 or 8.5% that men will have the final say in the decision. In this context it is expected that for men relative resources will not have an affect on women's marital power. For men in Sweden, the man having a higher relative education level increases the odds that women will have the marital power. A difference in higher relative education level increases the odds by 3.629 or 263% that women will make the final decisions. As seen in Table 4, as men age there is an increase in the odds that women will have power in the marital relationship. For each one year increase in men's age the odds of women making the decision increases by 1.030 or 3%. Philippines. In the Philippines it is expected that there will be no relationship between relative resources and man's or woman's power in the marital relationship. As shown in Table 4, there are no variables that significantly increase or decrease the odds that the men or women will have marital power. United States. It is expected in the United States that a significant relationship will be found between relative resources and marital power. However, for men only relative income has a significant relationship with marital power. This suggests that as men contribute less income there is a decrease in men's power. With a one unit increase in relative income there 42 is a decrease in the odds by 0.761 or 24% that the men will make the decisions. It is also found, as shown in Table 4, that the presence of children decreases the odds by 0.464 or 54% that men will make the decisions. Age significantly decreases the odds that men will have power in the marital relationship. With each one year increase in men's age there is a decrease in the odds that men will have final say in decision making by 0.974 or 3%. The finding for men regarding woman's marital power is that no relative resources increase or decrease the odds that women have power in the marital relationship. It is found however that family income significantly decreases the odds that the woman will have power in the marital relationship. As seen in Table 4, for decile increase of family income the odds decrease by 0.859 or 14% that women will make the decision. Women In this section the results for women in this investigation will be reported. It is expected that in Sweden and the Philippines, the egalitarian and traditional contexts respectively, no significant relationships will be found between relative resources and marital power. In the United States, the transitional context, is it expected that a significant relationship will be found between relative resources and marital power. All countries. As seen in Table 5, when women from all contexts are analyzed together relative resources do not significantly increase or decrease the odds of men's marital power. However it is seen in Table 5, that gender ideology significantly decreases the odds that men will make the decisions by 0.804 or 20%. This means as women's gender ideology becomes more egalitarian there is a significant decrease in the odds that men will have marital power. It is shown in Table 5, that age significantly decreases the odds that men will 43 make the decision by 0.997 or 0.03%. This means for each one year increase in women's age the odds decrease that men will have power in the marital relationship. Over all for women it is found that relative resources do not have a significant relationship with women's marital power. Gender ideology decreases the odds that women will have marital power. As gender ideology becomes more egalitarian there is a significant decrease in the odds that women will make the decisions by 0.825 or 17.5%. Over all for women it is seen that the presence of children increases the odds that women will make the decisions by 1.580 or 58%. Family income is also found to impact women's power in the marital relationship. Each decile increase of family income significantly lowers the odds that women will have final say in decisions by 0.947 or 5%. 44 Table 5 Female Participants Multinomial Logistic Regression All Countries Sweden USA Philippines Men's Power b Exp(B) b Exp(B) b Exp(B) b Exp(B) Intercept 0.636 0.883 0.131 -0.990 Lower Education 0.069 1.071 1.091 2.978 0.093 1.097 -0.217 0.805 Higher Education -0.280 0.756 0.965 2.625 -1.107" 0.331 -0.076 0.927 Relative Income -0.102 0.903 -0.045 0.956 0.045 1.046 -0.237* 0.789 Gender Ideology -0.218* 0.804 -0.604* 0.546 -0.209 0.812 0.279 1.322 Has Kids -0.384 0.681 -1.084 0.338 -0.191 0.826 -0.132 0.877 Family Income 0.002 1.002 0.042 1.043 -0.008 0.992 0.025 1.025 Age -0.023" 0.977 -0.033 0.968 0.015 0.985 -0.011 0.989 Women's Power Intercept 0.034 -1.223 -2.176* -0.518 Lower Education -0.291 0.748 -0.427 0.652 0.181 1.198 -0.360 0.698 Higher Education -0.035 0.966 -0.174 0.840 0.273 1.314 -0.092 0.912 Relative Income -0.068 0.935 0.102 1.107 0.093 1.098 -0.131* 0.877 Gender Ideology -0.193" 0.825 -0.079 0.924 0.269 1.308 -0.123 0.885 Has Kids 0.457" 1.580 0.471 1.601 0.151 1.163 0.251 1.285 Family Income -0.054* 0.947 -0.068 0.934 -0.074 0.929 -0.002 0.998 Age -0.005 0.995 -0.005 0.995 -0.004 0.996 0.014 1.014 *p = <0.05 **p = <0.01 "*p = < 0.001 #p = <0.1 Sweden. In the egalitarian context, Sweden, it is found that resources do not significantly increase or decrease the odds of the men's or women's power in the marital relationship. Gender ideology is the only variable with a significant impact on marital power for women and it is on men's power. It is seen in Table 5, as women's beliefs become more egalitarian there is a decrease in the odds that men will have the marital power. As woman's gender ideology becomes more egalitarian there is a decrease in the odds that men will make the decisions by 0.546 or 45%. PhUippines. In this context relative education level has no impact on the odds of men's or women's power in the marital relationship. The only variable for women that significantly impacts men's power, in this context, is relative income. Table 5 shows that relative income does significantly decrease the odds that men will make the decisions by 0.789 or 21%. For women in the Philippines, relative income is also the only variable that significantly impacts women's power. Relative income significantly decreases the odds that women will have power in the marital relationship. This means as women contribute more relative income to the marital dyad there is a decrease in the odds that women will make the decision by 0.877 or 12%. United States. In the transitional context of the United States it is expected for women that relative resources will have a significant relationship with men's and women's marital power. What is found for women is that only relative education matters for men's power. Table 5 shows that a difference in higher education level does significantly decrease the odds by 0.331 or 67% that men will make the decision. There are no variables for women in the United States that significantly increase or decrease the odds that women will have power in the marital relationship. 46 Modified Patriarchy It is hypothesized that there will be a difference between high and low income families for the relationship between relative income and marital power. The hypothesis is that high income families will transition to egalitarian beliefs about gender despite a strong culture norm of patriarchal gender ideology. Due to this gender ideology transition, it is believed that in high income families' relative resources will matter to attaining marital power. As seen in Table 6, the results for men, family income does appear to moderate the relationship between relative income and marital power but not in the manner that is expected. The findings indicate that for men there is a significant difference between high and low income families. The difference is that as women's income contribution increases there is a significant decrease in the odds that men in low income families will have final say in decisions by 0.766 or 23%. For the men there is no significant relationship found for women's power. 47 Table 6 Modified Patriarchy Results for Men in the Philippines Model I Model II Men's power b Exp(B) b Exp(B) Intercept -1.933 -1.271 Lower Education 0.252 1.286 0.223 1.250 Higher Education 0.210 1.233 0.268 1.308 Relative Income -0.004 0.996 0.088 7.093 Gender Ideology -0.093 0.911 -0.141 0.869 Has Kids 0.302 1.352 0.320 1.377 Family Income 0.014 1.014 -0.087 0.917 Age 0.010 1.010 0.010 1.010 Interaction -0.266* 0.766 Family Income X Relative Income Women's power Intercept -0.802 -0.572 Lower Education -0.023 0.977 0.422 1.525 Higher Education 0.416 1.516 -0.017 0.984 Relative Income 0.014 1.014 0.055 1.056 Gender Ideology -0.134 0.775 -0.153 0.858 Has Kids -0.255 1.007 -0.250 0.779 Family Income 0.007 1.014 -0.032 0.969 Age -0.002 0.998 -0.002 0.998 Interaction -0.097 0.908 Family Income X Relative Income *p = <0.05 **p = <0.01 ***p = < 0.001 #p = <0.1 As seen in Table 7, the results for women, family income does appear to moderate the relationship between relative income and marital power but not as expected. For women in low income families, as relative income contribution increases there is a significant increase in the odds by 1.40 or 40% that women will make the decision. In the women's findings family income does not moderate the relationship between relative resource contribution and marital power for men. 48 Table 7 Modified Patriarchy Results for Women in the Philippines Model I Model II Men's power b Exp(B) b Exp(B) Intercept -0.990 -1.495 Lower Education -0.217 0.805 -0.112 0.894 Higher Education -0.076 0.927 -0.272 0.762 Relative Income -0.237* 0.789 -0.397* 0.673 Gender Ideology 0.279 1.322 0.290 1.337 Has Kids -0.132 0.877 -0.098 0.906 Family Income 0.025 1.025 0.111 1.117 Age -0.011 0.989 -0.009 0.991 Interaction 0.263 1.301 Family Income X Relative Income Women's power Intercept -0.518 -1.183 Lower Education -0.360 0.698 -0.141 0.868 Higher Education -0.092 0.912 -0.434 0.648 Relative Income -0.131* 0.877 -0.348** 0.706 Gender Ideology -0.123 0.885 -0.103 0.902 Has Kids 0.251 1.285 0.283 1.327 Family Income -0.002 0.998 0.113* 1.120 Age 0.014 1.014 0.016 1.016 Interaction 0.336** 1.400 Family Income X Relative Income *p = <0.05 **p = <0.01 ***p = < 0.001 #p = <0.1 49 Discussion The purpose of this investigation is to answer the question: Do resources matter to the acquisition of power in marital relationships? This study found that the resources do impact power acquisition, but not as expected. Blood and Wolfe (1960) suggest that as relative resource contribution increases, the marital power of the contributor will also increase. Yet, this is not what this study's findings support. In this study the findings are that resources do not directly increase the participant's power. Rather the findings show that resource contribution is perceived to affect the partner's power. In this investigation it is seen that as women contribute greater relative resources to the marriage the odds decrease that their spouse will have the marital power. Yet, when men are found to have a greater relative resource contribution, women's power increases. In this investigation, country context as a moderator of the relationship between resources and marital power is also of interest. Country context does not have the expected affect on the relationship between resources and marital power. Rodman (1972) suggests that resource contribution will only matter in the transitional context because the norms about gender are weak, uncertain, and flexible. Without strong social norms and expectations about how power should be divided between men and women, Rodman hypothesizes that the individual in the transitional context who contributes the most valued and needed resources will acquire the most marital power. Rodman's hypothesis of resources only affecting power acquisition in the transitional context was not supported in this investigation. In this study neither men nor women in the transitional context experience greater odds of acquiring marital power with greater resource contribution in the marital relationship. In the egalitarian and traditional context resource 50 contribution was not expected to impact marital power. Yet, relative education is found to affect the odds of marital power in the egalitarian context and relative income is found to impact marital power in the traditional context. The concept of separate spheres may explain some of the findings in this study. According to Laslett and Brenner (1989), separate spheres have historically been central to the social organization of gender. The idea of separate spheres is a manner of thinking about gender that dichotomizes men and women into distinct areas of dominance. Traditionally, women have been seen as responsible for the domestic domain and household labour, while men have been seen as responsible for providing financially for the family by participating in the public workplace (Runte & Mills, 2004). One explanation for why women are found to possess marital power more often than men could be that the power variable used in this study may simply consist of domestic domain decisions. The belief that the domestic sphere is a woman's domain, especially in the patriarchal context, could explain the number of respondents in this study who report women as having the power (Laslett & Brenner, 1989; Runte & Mills, 2004). It could be that what is measured in this study is not marital power but rather the level of separateness between the domestic and work spheres in each country. Sweden, where men and women are expected to equally participate in domestic and work domains, could be an example of a country where there is high overlap between the spheres. This is demonstrated by the number of respondents in Sweden that report shared decision making and the low number that report women have the power. Support for this idea also is found in the Philippines, where power is indicated to be shared the least, and a higher number of respondents report that women possess the marital power. The findings suggest that the 51 traditional context has the most segregated work and domestic domains. The spheres being highly segregated in the Philippines indicates that women are expected to complete all of the household tasks and men are not expected to participate in the domestic sphere. This supports the idea that the decision making indicators in this investigation may not be reflective of marital power. Instead of measuring marital power the power variable may actually measure expectations about sphere participation and the degree of separateness between the domestic and work spheres. A quick assessment of shared power could lead to an assumption of equality in marriage. It is intuitive that the egalitarian country, with the highest report of shared power is therefore is the most equal and the traditional country with the lowest report of shared power is the least equal. Yet, despite beliefs of gender equality in Sweden, women still complete the majority of household labour (Nordenmark & Nyman, 2003). Inequality in household labour suggests that Sweden is not as equal in practice despite having the most egalitarian beliefs. This indicates that there is a possibility that shared power is not an indicator of equality, but perhaps there is another interpretation of the finding of shared power by the participants. It is possible that shared power is not gender equality, but rather an indication that there is a negotiation process that precedes decisions in marriage. These findings indicate then that the most negotiation about decisions occurs in Sweden, and the least amount of negotiation happens in the Philippines. Negotiation in the power literature is referred to as power process (Cromwell & Olson, 1975a). As mentioned earlier, the inability to capture power process has been one of the largest criticisms of decision making measures. If shared power is an indication of negotiation, it could be that the participants are indicating a similar 52 consultation process to that which Blood and Wolfe (1960) assumed was present in marriage. However contrary to Blood and Wolfe's (1960) consultation process that leads to one spouse having the power to make the decision, it may be that couples are indicating a more intensive joint negotiation process where ultimately both members of the marital dyad share responsibility for any decisions. Possibly the respondents do not see negotiation as a manner of influencing and exerting power over their partner (Safilios-Rothschild, 1970). Rather, than the man or woman holding power individually, shared power may represent negotiation that leads to compromise and cooperation in the marital relationship. As previously mentioned, final say decision making measures have been heavily critiqued in the marital power literature. Two of these critiques apply to this research investigation. The first critique is that the power variable may not be gender neutral. Gender neutrality is a requirement that Blood and Wolfe (1960) specify as necessary for any decision to be used in a marital power measure. For example, the lack of specificity of what a major purchase for the home is leaves respondents having to make their own assumptions about what type of purchase is a home purchase. If the respondent imagined the purchase to be gendered, it may have led to a response of male or female as the decision maker. Rather than gender neutrality, the respondent's interpretation of the item could influence the response regarding who has marital power. Being that the purchase was for the home, may have led to more women being indicated as having marital power, than if the indicator had specified that the decision was about a major purchase for the couple. The next critique of final say decision making measures is that the person indicated as the decision maker may simply be the person who is told to decide rather than the individual with marital power (Centers et al, 1971; Safilios-Rothchild, 1970). Delegation is believed to 53 occur when a decision is considered to be unimportant. This is one of the issues that Safilios-Rothschild (1970) discusses as a major drawback to assessing marital power with decision making measures. It may be that women do not have more power than men in the Philippines. It could simply be that the indicators of marital power in this study are considered unimportant decisions and have therefore been placed under female jurisdiction. The concept of delegation suggests, that Filipino women may be inaccurately classified as having marital power in this study. The truly powerful person in the relationship is the person who is able to delegate the decisions (Safilios-Rothchild, 1970). Final say decisions making indicators, may have more value in one country over another. For example, it is seen that in the United States that men are reported as having marital power more often than women after shared power. There could be a cultural difference in how important these decisions are across countries. This difference may be what is reflected in the findings. In the United States where possessions indicate wealth and status, major purchases for the home may be considered important decisions. When decisions are considered to be important, it is possible that there is a greater likelihood of men making the decision. This may explain why in the United States men are reported to have marital power more often than women. In the Philippines, major purchases for the home may not be seen as important and are therefore delegated to women. This suggests that the findings reflect attitudes and cultural beliefs about the actual decisions, rather than the decisions simply being neutral indicators of marital power. Using different resources in this study could have led to different findings regarding the relationship between relative resources and marital power. This investigation relied solely on external and tangible resources. Other resources such as internal resources might 54 offer a better explanation of what increases or decreases the odds of power in the marital relationship. Internal resources that are suggested to have an impact on marital power are emotional support and maintenance of social networks (Heer, 1963; Scanzoni, 1979; Szinovacz, 1986). Safilios-Rothschild (1970) suggests internal resources are more difficult to measure and evaluate in regards to marital power. However, if the power that is being measured does fall in the domestic sphere it is possible that internal resources may have been more appropriate for this investigation. Children may be a potential resource for women in relationships. Traditionally children have been seen as a negative factor for women in marital relationships. Blood and Wolfe (1960) found that the presence of children decreased women's marital power. The findings of this study show that children decrease men's power and increase women's power. This suggests that women may experience greater marital power when children are present in the household. This may be a new manner of conceptualizing children for marital power research. If children are a resource, the ability to contribute children to the marriage may offer women an opportunity to gain marital power. It is also possible that since children are seen as residing in the domestic sphere, children may be capital that women use to decrease men's power. The idea of children as a resource offers a different perspective on the presence of children for future marital power research. Resources may not be the main influence on marital power. There are many other aspects internal and external to the marital relationship that may impact how power is structured. Personal characteristics such as perceived competency, knowledge, or ability are suggested to impact which partner has the final say in specific decisions (Bahr, 1972; Blood & Heer, 1963; Brinkerhoff & Lupri, 1989; Cromwell & Olson, 1975a, 1975b; Safilios-55 Rothchild, 1970; Scanzoni, 1979). Relationship characteristics could also impact who has marital power. Health status of both members of the relationship could impact who is the final decision maker and has power. If one spouse has a disability that prevents sound decision making, the other spouse may make the decisions and therefore be assessed as having the power. Marital status is also a characteristic of the relationship that may impact the power dynamic. When couples are living common law, it is possible that more value is placed on equality and a 50/50 split in all areas of the relationship. Power could be a very important aspect of maintaining the 50/50 nature of the relationship. Once couples are married, equity or a feeling of fairness, may become more important than equality (Thompson, 1991). Due to a belief in equity, married individuals may be more interested in fairness or sharing power than in having greater marital power than their partner. Subtleties such as equality versus equity in married and common law couples can not be isolated in this investigation, since the 2002 ISSP does not identify married and common law individuals separately. Before discussing the impact of gender ideology, it is important to recognize that other factors may influence the relationship between resources and marital power. It is possible that beliefs and values, such as political and religious freedom may impact power structures between men and women in marriage. Countries that are more politically and religiously restricted may have stronger values and beliefs about how power should be divided between genders. Understanding the political and social climate could be an important aspect to consider when completing marital power research, and these aspects are not captured in this study. 56 In this investigation, the context is described as values and beliefs about gender. It is believed that these cultural values and beliefs about gender specify how power should be divided between men and women. Using a value based definition of gender ideology, as the cultural context, this study aims to understand how individuals act differently in similar situations across contexts (Swidler, 1986). This suggests that culture is based on societal values and therefore values are the reason that individuals act and react in certain ways to specific situations (Swidler, 1986). It could be that the gender ideology scale only captured a small portion of attitudes and beliefs about gender. In this investigation the only indicators used to measure gender ideology are participants' attitudes towards gender in the domestic and work sphere. There is more to gender ideology than simply attitudes towards working outside the home. Conceptualization of gender as an unspoken and unconscious way of being in the world, the idea of doing gender (Berk, 1985), is an important part of gender ideology that is not captured in this study. The subtleties of gender ideology can not be accessed due to a lack of indicators in the data set. The use of a gender ideology scale is important, rather than simply designating each country a gender ideology based on supposition and arbitrary assignment. Yet, there is a need to continue improving the indicators of gender ideology to ensure a better understanding of the country context. Swidler suggests that rather than using values to connect culture to action, perhaps culture is a "tool kit" of possible actions to be taken (Swidler, 1986, p. 273). This tool kit of culture holds the rituals, symbols, skills, and habits that capitalize on the strengths and abilities of a society (Swidler, 1986). These tools according to Swidler (1986) are not a unilateral set of values about gender. When faced with situations individuals will access this 57 metaphorical took kit and gather possible ideas for action. Rather than a one set of specific values per country it is possible that in each country, individuals simply access different tools in the tool kit. Applying Swidler's ideas to this investigation, shows that the gender ideology scale may be limited within this research study as it is operationally defines gender ideology as a set of values and beliefs that each individual holds about gender. Swidler's idea suggests that a theoretical shift may be necessary to create a multidimensional view of gender ideology as the cultural context. This shift may allow for individuals to capitalize on the tools inherent in each society, such as "culturally based skills, habits and styles" (Swidler, 1986, p. 275) that promote growth rather than a rigid set of beliefs and values. Future Research When discussing future research, identifying the challenges present in this study that should be addressed in future research is important. The first improvement is to use more than two decision making variables to measure power. Blood and Wolfe's original scale had eight different decision making questions designed to target major areas of decision making for couples (Allen & Straus, 1984; Blood & Heer, 1963; Blood & Wolfe, 1960; Brinkerhoff & Lupri, 1989; Diefenbach, 2002; Elliott & Moskoff, 1983; Rodman, 1967, 1972; Safilios-Rothchild, 1967, 1970). Only targeting two areas of decision making, may limit the ability of this study to make generalizations about the relationship between resources and power. Another challenge to be addressed in future studies is the use of an individual level of analysis. In this investigation only one member of the marital dyad reports on the power in the relationship either the husband or the wife. Historically only wives have been targeted to offer information about the power relationship and this is believed to have biased the 58 generalizations made about marital power (Allen & Straus, 1984; Brinkerhoff & Lupri, 1989; Burr et al., 1977; Cromwell & Olson, 1975a; McDonald, 1980; Safilios-Rothchild, 1970; Scanzoni, 1979). Not limiting this investigation to only women begins to develop a multi-gendered perspective of outcome power. Measuring both members of the marital dyad would more fully address this issue in the research. Investigating both husbands and wives in future research will allow for a direct comparison of power in marriage. By measuring the dyad it is possible to begin to understand some of the different perspectives that men and women may report about the same marriage. Using only opposite sex couples in this research limits the view of marital power. In the future using same sex couples will add to the present body of marital power literature. Although same sex marriage was not legal in 2002 it is possible that the respondents who were same sex and living with their partners may have responded as living as married. The 2002 ISSP did not differentiate between living as married opposite sex and same sex couples and therefore in this investigation all couples were treated as opposite sex couples. However in the future, with the growing acceptance of same sex couples and the legalization of same sex marriage it seems that there will be a greater opportunity to include sex couples in marital power research. It is also possible that the inclusion of same sex couples in marital power research may lead to thinking beyond gender and gender ideology to try to understand power structure in marital relationships. Future research should also continue to test the impact of context on marital power in marriage. This future research should be focused on two separate areas. The first area should be on the statistical analysis used to measure the moderating affect of cultural context. By including more countries in the analysis it is possible to use Hierarchal Linear Modeling 59 (HLM) to test the strength of context as a moderator. HLM provides another method of analysis that offers a more robust test and allows for greater cross cultural comparison to be completed. The second area is to create measures based on Swidler's ideas of context. It is possible in future research there may be creative ideas and indicators that could capture the idea of culture as a tool kit. These different indicators may allow investigators to limit the use of values and beliefs that could unintentionally offer an incomplete and unilateral view of culture. Finally one of the most interesting future directions to pursue is to develop new measures of marital power. By creating new power measures for survey research it will be possible to address the numerous criticisms offered by Safilios-Rothschild (1970) and other authors directed at final say decision making indicators. The current decision making measures allow power research to occur cross culturally. However, new measures might revitalize an interest in marital power. It is possible that decision making may not be used to evaluate power in future research, but rather there may be other indicators that could capture the marital power as a dynamic relational concept. However, these new measures have not yet been conceptualized. For the past 45 years marital power has been heavily dominated by final say decision making, a shift in thinking is necessary to begin a new era of marital power research. In summation marital power is a dynamic and important concept to investigate. Equality in power in marital relationships is connected to individual physical and mental well being. Past research has attempted to explain and understand marital power by focusing on individual's contributions to the marital relationship. One such theory offered by Blood and 60 Wolfe (1960) has influenced marital power research for the past 45 years. Rodman's (1972) theory of resources in a cultural context takes a unique perspective on marital power. Rodman's theory contextualizes the relationship between resources and marital power by adding the impact of cultural context. In this investigation relative resources do not have the expected relationship with marital power when investigating individuals. It is also found that when testing for the moderating affect of context the findings were not as initially expected. Despite the findings in this study, it is still believed that context is a powerful moderator that needs to be included in research investigations. It is possible that the test offered by this investigation is limited due to the conceptualization and operationationalization of the indicators for both resources and marital power. In the future context will continue to be tested to understand the impact that it may have other power relationships. 61 References Adams, B. N. (2004). Families and family study in international perspective. Journal of Marriage and Family, 66(4), 1076-1088. Agarwal, B. (1997). "Bargaining" and gender relations: Within and beyond the household. Feminist Economics, 3(1), 1-51. Allen, C. M. (1984). On the validity of relative validity studies of "final-say" measures of marital power. Journal of Marriage and the Family, 46(3), 619-629. Allen, C. M., & Straus, M. A. (1984). "Final say" measures of marital power: Theoretical critique and empirical findings from five studies in the United States and India. Journal of Comparative Family Studies, XV(3), 329-344. Alverez, M. (1979). Family power structure in Chile: A survey of couples with children in primary schools. International Journal of Sociology of the Family, 9, 123-131. Amato, P. R., Johnson, D. R., Booth, A., & Rogers, S. J. (2003). Continuity and change in marital quality between 1980 and 2000. Journal of Marriage and Family, 65(1), 1-22. Bahr, S. (1972). Comment on "the study of family power structure: A review 1960-1969". Journal of Marriage and the Family, 34(2), 239-243. Berk, S. F. (1985). The gender factory: The apportionment of work in american households. New York: Plenum. Blood, R. O., & Heer, D. (1963). The measurement and bases of family power: A rejoinder. Marriage and Family Living, 25(4), 475-478. Blood, R. O., & Wolfe, D. M. (1960). Husbands & wives: The dynamics of married living. Illinois: The Free Press of Glencoe. 62 Bolak, H. C. (1995). Towards a conceptualization of marital power dynamics: Women breadwinners and working-class households in turkey. In S. Tekeli (Ed.), Women in a modern turkish society (pp. 173-198). New Jersey: Zed Books Ltd. Brinkerhoff, M. B., & Lupri, E. (1989). Power and authority in the family. In K. Ishwaran (Ed.), Family and marriage: Cross cultural perspectives. Toronto: Wall and Thompson. Burr, W. R., Ahern, L., & Knowles, E. M. (1977). An empirical test of Rodman's theory of resources in a cultural context. Journal of Marriage and the Family, 39(3), 505-514. Centers, R., Raven, B. H., & Rodrigues, A. (1971). Conjugal power structure: A re examination. American Sociological Review, 36(2), 264-278. Conklin, G. H. (1979). Cultural determinants of power for women within the family: A neglected aspect of family research. Journal of Comparative Family Studies, X(\), 35-53. Connnell, R. W. (1987). Gender regimes and the gender order. In Gender and power: Society, the person and sexual politics. Stanford: Stanford University Press. Corrales, R. G. (1975). Power and satisfaction in early marriage. In R. E. Cromwell & D. H. Olson (Eds.), Power in families (pp. 197-217). Toronto: John Wiley & Sons. Cromwell, R. E., & Olson, D. H. (1975a). Methodological issues in family power. In R. E. Cromwell & D. H. Olson (Eds.), Power in families (pp. 131-151). Toronto: John Wiley & Sons. Cromwell, R. E., & Olson, D. H. (1975b). Multidisciplinary perspectives of power. In R. E. Cromwell & D. H. Olson (Eds.), Power in families (pp. 15-37). Toronto: John Wiley & Sons. 63 Dhruvarajan, V. (1992). Conjugal power among first generation hindu asian indians in a Canadian city. International Journal of Sociology of the Family, 22(spring), 1-34. Diefenbach, H. (2002). Gender ideologies, relative resources, and the division of housework in intimate relationships: A test of hymen Rodman's theory of resources in a cultural context. International Journal of Comparative Sociology, 43(1), 45-64. Elliott, J. E., & Moskoff, W. (1983). Decision-making power in romanian families. Journal of Comparative Family Studies, XIV(l), 37-50. Felmlee, D. (1994). Who's on top? Power in romantic relationships. Sex Roles Retrieved February 26, 2005, from http://www.fmdarticles.eom/p/articles/mi m2294/is n5-6 v31/ai 16498784/print Godwin, D. D., & Scanzoni, J. (1989). Couple consensus during marital joint decision making: A context, process, outcome model. Journal of Marriage and the Family, 51(4), 943-956. Goldhamer, G., & Shils, E. A. (1939). Types of power and status. The American Journal of Sociology, 45(2), 171-182. Gray-Little, B. (1982). Marital quality and power processes among black couples. Journal of Marriage and the Family, 44(3), 633-646. Greenstein, T. N. (1996a). Husbands' participation in domestic labor: Interactive effects of wives' and husbands' gender ideologies. Journal of Marriage and the Family, 58(3), 585-595. Greenstein, T. N. (1996b). Gender ideology and perceptions of the fairness of the division of household labor: Effects on marital quality. Social Forces, 74(3), 1029-1042. 64 Grinberg-Vinaver, S. (1955). The status of women throughout the world. Marriage and Family Living, 17(3), 197-204. Hallenbeck, P. N. (1966). An analysis of power dynamics in marriage. Journal of Marriage and the Family, 28(2), 200-203. Heer, D. (1963). The measurement and bases of family power: An overview. Marriage and Family Living, 25(2), 133-139. Hesse-Biber, S., & Williamson, J. (1984). Resource theory and power in families: Life cycle considerations. Family Process, 23, 261-278. Kohn, M. L. (1987). Cross- national research as an analytic strategy: American sociological association, 1987 presidential address. American Sociological Review, 52(6), 713-731. Komter, A. (1989). Hidden power in marriage. Gender & Society, 3(2), 187-216. Laslett, B., & Brenner, J. (1989). Gender and social reproduction: Historical perspectives. Annual Review of Sociology, 15, 381-404. Lavee, Y., & Katz, R. (2002). Division of labour perceived fairness, and marital quality: The effect of gender ideology. Journal of Marriage and Family, 24(1), 27-39. Lee, G. R., & Petersen, L. R. (1983). Conjugal power and spousal resources in patriarchal cultures. Journal of Comparative Family Studies, XVI(l), 23-31. McDonald, G. W. (1980). Family power: The assessment of a decade of theory and research. Journal of Marriage and the Family, 42(4), 841-854. Min, G. P. (1997). Korean immigrant wives' labor force participation, marital power and status. In E. Higginbotham & M. Romero (Eds.), Women and work: Exploring race, ethnicity and class (Vol. 6, pp. 176-191). Thousand Oaks: Sage Publications. 65 Mizan, A. (1994). Family power studies: Some major methodological issues. International Journal of Sociology of the Family, 24(2), 85-91. Nordenmark, M. (2004). Does gender ideology explain differences between countries regarding the involvement of women and men in paid and unpaid work? International Journal of Social Welfare, 13, 233-243. Nordenmark, M., & Nyman, C. (2003). Fair or unfair? Perceived fairness of household division of labour and gender equality among men and women. The European Journal of Women's Studies, 10(2), 181-209. O'Connor, P. (1991). Women's experience of power within marriage: An inexplicable phenomenon. Sociological Review, 39(4), 823-842. Olson, D. H., & Cromwell, R. E. (1975). Power in families. In R. E. Cromwell & D. H. Olson (Eds.), Power in families (pp. 3-14). Toronto: John Wiley & Sons. Olson, D. H., & Rabunsky, C. (1972). Validity of four measures of family power. Journal of Marriage and the Family, 34(2), 224-234. Oropesa, R. S. (1997). Development and marital power in Mexico. Social Forces, 75(4), 1291-1315. Rodman, H. (1967). Marital power in France, Greece, Yugoslavia, and the United States: A cross-national discussion. Journal of Marriage and the Family, 29(2), 320-324. Rodman, H. (1972). Marital power and the theory of resources in a cultural context. Journal of Comparative Family Studies, 3, 50-69. Rollins, B. C, & Bahr, S. (1976). A theory of power relationships in marriage. Journal of Marriage and the Family, 38(4), 619-627. Ross, C. E. (1987). The division of labour at home. Social Forces, 65(3), 816-833. 66 Runte, M., & Mills, A. J. (2004). Paying the toll: A feminist post-structural critique of the discourse bridging work and family. Culture and Organization, 10(3), 237-249. Sabatelli, R. M., & Shehan, C. L. (1993). Exchange and resource theories. In P. G. Boss, W. J. Doherty, R. La Rossa, W. R. Schumm & S. K. Steinmetz (Eds.), Sourcebook of family theories and methods. New York: Plenum Press. Safilios-Rothchild, C. (1967). A comparison of power structure and marital satisfaction in urban greek and french families. Journal of Marriage and the Family, 29(2), 345-352. Safilios-Rothchild, C. (1970). The study of family power structure: A review 1960-1969. Journal of Marriage and the Family, 32(4), 539-552. Scanzoni, J. (1979). Social process and power in families. In W. R. Burr, R. Hill, F. I. Nye & I. L. Reiss (Eds.), Contemporary theories about the family (Vol. 1, pp. 295-316). New York: The Free Press. Sprecher, S., & Felmlee, D. (1997). The balance of power in romantic heterosexual couples over time from "his" and "her" perspectives. Sex Roles, 37(5/6), 361-379. Sprey, J. (1972). Family power structure: A critical comment. Journal of Marriage and the Family, 34(2), 235-238. Sprey, J. (1975). Family power and process: Toward a conceptual integration. In R. E. Cromwell & D. H. Olson (Eds.), Power in families (pp. 61-80). Toronto: John Wiley & Sons. Sprey, J. (1999). Family dynamics: An essay on conflict and power. In M. B. Sussman, S. K. Steinmetz & G. W. Peterson (Eds.), Handbook of marriage and the family (2nd ed., pp. 667-685). New York: Plenum Press. 67 Swidler, A. (1986). Culture in action" symbols and strategies. American Sociological Review, 51, 273-286. Szinovacz, M. (1986). Family power. In M. B. Sussman & S. K. Steinmetz (Eds.), Handbook of marriage and the family (pp. 651-693). New York: Plenum Press. Thompson, L. (1991). Family work: Women's sense of fairness. Journal of Family Issues, 12(2), 181-196. Warner, R. L., Lee, G. R., & Lee, J. (1986). Social organization, spousal resources, and marital power: A cross cultural study. Journal of Marriage and the Family, 48(\), 121-128. Weber, M. (1947). The theory of social and economic organization. New York: Oxford University Press. Webster, C. (2000). Is spousal decision making a culturally situated phenomenon? Psychology & Marketing, 77(12), 1035-1058. Xu, X., & Lai, S. (2002). Resources, gender ideologies, and marital power: The case of Taiwan. Journal of Family Issues, 23(2), 209-245. Yodanis, C. L. (2003). Power: Marital relationships. In J. J. Ponzetti (Ed.), International encyclopedia of marriage and family (2nd ed.). New York: Macmillan Reference USA. Zipp, J. F., Prohaska, A., & Bemiller, M. (2004). Wives, husbands and hidden power in marriage. Journal of Family Issues, 25(7), 933-958. Zuo, J. (2004). Shifting the breadwinning boundary: The role of men's breadwinner status and their gender ideologies. Journal of Family Issues, 25(6), 811-832. 68 Appendix Appendix A Aggregate Mean Gender Ideology Country 1988 1994 2002 Austria 2.43 2.67 2.90 Australia 3.03 3.01 Bulgaria 2.65 Brazil 2.29 Chile 2.46 Cyprus 3.02 Czech Republic 2.72 2.72 Denmark 3.63 Finland 3.25 Flanders (Belgium) 2.87 France 3.08 Great Britain 3.11 3.20 3.19 Hungary 2.29 2.20 2.51 Ireland 2.76 2.90 3.15 Israel 2.81 3.25 3.15 Japan 2.73 2.86 Latvia 2.77 Mexico 2.48 Netherlands 3.06 3.18 3.14 New Zealand 3.18 3.12 Northern Ireland 3.07 3.08 Norway 3.32 3.51 Philippines 2.39 2.40 Poland 2.48 2.75 Portugal 2.75 Russia 2.13 2.56 Slovenia 2.69 2.77 Slovakian Republic 2.49 Spain 2.91 3.04 Sweden 3.34 3.39 Switzerland 2.90 Taiwan 2.63 United States 3.06 3.11 3.09 69 

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 11 0
China 3 5
Japan 2 0
Kenya 1 0
Canada 1 1
Germany 1 0
City Views Downloads
Ashburn 4 0
Washington 4 0
Beijing 3 0
Tokyo 2 0
Sunnyvale 1 0
Princeton 1 0
Redmond 1 0
Unknown 1 0
Calgary 1 1
Nairobi 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092699/manifest

Comment

Related Items