Sex Roles (2009) 60:843–858 DOI 10.1007/s11199-009-9591-y

ORIGINAL ARTICLE

Political Distrust and Gendered Attitudes: The Japanese State and Women Kana Fuse & Nanaho Hanada

Published online: 27 February 2009 # Springer Science + Business Media, LLC 2009

Abstract We explore sources of political trust by examining one’s gender role attitudes as they relate to political trust in Japan. We hypothesize that progressive gender role attitudes are associated with political distrust. This political distrust is perhaps caused by relative deprivation which progressive individuals experience, given that gender relations are still traditional in Japan. We further hypothesize that this distrust among women extends to the government itself. We analyze the 2003 Japanese General Social Survey, a national representative survey of Japanese adults. Based on multivariate analyses of 1,807 respondents, we find support for our hypotheses. Our findings suggest that gender issues are salient among progressive individuals, especially among women. Keywords Gender role attitudes . Political trust . Women . Relative deprivation . Japan Introduction “Because the number of childbearing machines and devices is fixed, all we can ask is for each woman to do their best job.” —Hakuo Yanagisawa, Minister of Health, Labor and Welfare of Japan, 2007 (“Japan Minister Rapped on Women,” Jan 29th, 2007) K. Fuse (*) Institute for Health Metrics and Evaluation, University of Washington, 2301 5th Ave. Suite 600, Seattle, WA 98121, USA e-mail: [email protected] N. Hanada Department of Political Science, The Ohio State University, 2014 Derby Hall, 154 N. Oval Mall, Columbus, OH 43210, USA

“It is truly strange to say that we have to use tax money to take care of women who don’t even give birth once, who grow old living their lives selfishly.” —Yoshiro Mori, Ex-Prime Minister of Japan, 2003 (Beech 2005) In this study, we explore whether gender role attitudes have an independent effect on voters’ trust in political leaders and the government in Japan. Studying political trust has both empirical and normative implications. First, by finding sources of political trust, we may be able to increase trust. Decline in political trust has been paralleled by declining voter turnout across advanced democracies such as the U.S. (Hibbing and Theiss-Morse 1995) and in Britain, Canada, Finland, Germany, Iceland, Ireland, Japan, Norway, and Sweden (Pharr and Putnam 2000a). As Pharr and Putnam’s edited volume shows, Japan is not an exception as voter turnout in national general elections has decreased in the post war era. The 1994 electoral law changed from a single-non-transferable vote to a mixed system of single-member-districts and proportional representation to elect the lower house of the national parliament. Although proportional representation systems have been found to have a higher voter turnout than singlemember-district systems, (Endersby and Kriechkaus 2008; Powell 1986), Japan faced its lowest turnout rate in the post WWII history at the first general elections after the electoral law change. Only 59.65% and 59.62% of registered voters in Japan casted their votes for single-member district and proportional representation ballots, respectively. Therefore, identifying sources of political trust may help offset the declining voter turnout in the long run. Additionally, because political trust captures the underlying support for the political system, it is critical for democratic stability.

844

Sex Roles (2009) 60:843–858

Empirically, Japan is an interesting case. Although Japan is one of the largest economies in the world, Japanese women are socially, economically, and politically subordinate. While the Japanese are becoming more educated and progressive, women’s disadvantaged situation has barely improved. This is also reflected in recent sexist comments made by prominent politicians (see quotes above). We suggest that distrust of political leaders among the Japanese may be attributed to relative deprivation (Gurr 1968) experienced by liberal individuals living in a society with traditional gender relations. ‘Relative deprivation’ is defined as a gap between what one believes s/he deserves and what s/he actually gets (Gurr 1968). This gap can be created by comparing one’s current situation to: (1) others; (2) oneself in the past; or (3) one’s attitudes, preferences, and expectations (for further discussion, see Walker and Smith 2002). In this paper, we are interested in the gap between the current state of gender relations and one’s gender role attitudes and expectations. Because women are the ones who are directly affected by the nature of societal gender relations, we expect that the target of distrust extends to the government itself among progressive women. This notion will be discussed in further detail in the following section of this paper. This study’s empirical findings are relevant not just for Japan, but all consolidated democracies and democratizing countries. First, as mentioned earlier, a decline in political trust is observed in advanced democracies including the U.S. (Hibbing and Theiss-Morse 1995), Britain, Canada, Finland, Germany, Iceland, Ireland, Japan, Norway, and Sweden (Pharr and Putnam 2000b). Provided that a decline in political trust accompanies a similar decline in voter turnout, sources of political trust should be investigated. Second, in the past few decades, more countries have begun to democratize. For those countries, generating and maintaining a reasonable level of political trust among citizens is imperative for consolidation of democracy. However, in some countries such as post-communist regimes, democratization has been accompanied by a decline in women’s status (LaFont 2001). Therefore, if stagnation in gender relations leads to dissipated political trust among citizens, it will send a critical message to policy makers in many parts of the world.

the belief that the public cannot influence political outcomes because government leaders and institutions are unresponsive” (Miller et al. 1980, p 253). When voters are dissatisfied with the performance of current office holders, they can simply vote them out in elections. Generally, dissatisfaction with current leaders does not translate into dissatisfaction with the system itself (Dalton 2006). This is because when political power shifts hands frequently, the concept of the system can be distinguished from that of incumbent office holders. However, extending the distrust of incumbents to a diffused target (such as the government as an institution) is likely in countries like Japan. In Japan, the center-right Liberal Democratic Party (LDP) has been in power continuously since 1955, except for a short period in 1994 when Japan was ruled by a grand coalition formed by all of the opposition parties excluding the Communist Party. Therefore, the LDP’s performance may be reflected in the evaluation of the system itself when dissatisfaction is more imminent. In short, in the case of Japan, we argue that voter’s dissatisfaction with specific administrations (specific support) can be extended to a distrust of institutions and the system itself beyond specific office holders (diffused support) if this dissatisfaction felt by the voters is imminent. Scholars studying the American case have argued that women and men have different political interests where women are more likely to care about welfare, environment, childcare, and gender related issues (Deitch 1988; Page and Shapiro 1992; Wirls 1986). While there may be gender differences in the impact of one’s gender role attitudes on political trust, the literature on political trust has not focused on gender differences. Moreover, studies of the impact of incumbent performance on trust in the U.S. are limited to economic issues (Citrin and Green 1986; Miller 1983; Weatherford 1984, 1987), race and Vietnam issues (Miller 1974), and personal characteristics of politicians (Citrin 1974; Citrin and Green 1986). Government outputs are not limited to these areas, however. Our goal in this paper is to assess whether, and to what extent, the government’s performance in gender issues is reflected in the political trust of the people. Gender equality has not been a major issue in electoral campaigns in Japan. Are political parties missing potential mobilization?

Political Trust

Japanese Women

According to Easton (1965, 1975), political trust has two levels. One is a belief in the responsiveness of the current administration (specific support). The other is a belief in the responsiveness of the system itself, beyond each administration (diffused support, i.e. electoral procedures; Anderson and Guillory 1997). Political distrust “indicates

Although the Japanese are known to hold more traditional gender role attitudes compared to other industrialized countries (Treas and Widmer 2000), it has been observed that they have become more progressive in gender role views over the past few decades. A survey of single men and women in their reproductive years conducted every

Sex Roles (2009) 60:843–858

5 years shows that the percentage of people who agree to the statement, “men should work outside and women should stay home after marriage” has declined considerably. This percentage has dropped from 51% and 62% (women and men, respectively) in 1992 to 29% and 36% in 2005 (National Institute of Population and Social Security Research 2007). The Japanese are also becoming more educated. The percentage of women entering 4-year universities has steadily increased from 15% in 1990 to 32% in 2003, and among men, it has increased from 34% to 41% (OECD 2004). Although entering a junior college (2-year college) as opposed to a 4-year university used to be a popular choice among women, today Japanese women are shying away from junior colleges. In 1990, 22.2% of women attended junior colleges but the percentage has dropped to 13.9% in 2003 (OECD 2004). Moreover, there has been an increase in the percentage of the population pursuing graduate degrees. The percentage of Japanese attending graduate schools has increased from 3.1% in 1990 to 6.8% in 2003 among women and from 7.7% in 1990 to 13.8% in 2003 among men (Cabinet Office of Japan 2006). Despite the fact that the Japanese are becoming more progressive in their gender role attitudes and are becoming more educated, Japan still lags in terms of gender relations in comparison to other industrialized countries. For instance, Japanese husbands are reported to perform very little household labor (Davis and Greenstein 2004). The gender pay gap is also larger than in most developed countries (e.g. 66.0% in Japan; 76.4% in the U.S.; 91.2% in Sweden). In the Japanese labor force, only 10% of managerial positions were held by women in 2005, compared to 38, 32, and 42% in the same year in Germany, Sweden, and the U.S., respectively (United Nations Statistics Division 2007). Moreover, only 10% of tenured professors in Japan were women in 2004 (Cabinet Office of Japan 2006). In contrast, the percentages of tenured professors who were women were 26% in Canada in 2003 (Canadian Association of University Teachers 2004) and 38% in the U.S. in 2001 (U.S. Department of Education 2003). Women’s advancement in the political sphere is also minimal. The representation of women in the legislative branches at both national and local levels is very low. As of November 2006, 9.4% of seats in the Lower House of the national parliament were held by women, and as of December 2005, only 8.8% of seats in local assemblies were held by women. Examples of percentages of women in the National parliament in other developed countries are 24.7% in Australia, 31.6% in Germany, and 16.3% in the U.S. (Inter-Parliamentary Union 2007). With an increase in the number of progressive and educated women, it is not

845

surprising that Japanese women perceive men are as having a greater advantage socially and politically. In 2002, the majority of women reported they believe that they are at a disadvantage in employment (67.5%) and in the political sphere (82.9%) (Cabinet Office of Japan 2002). These facts suggest possible relative deprivation among progressive individuals because of the gap between their expectations and their surrounding reality. Gender Role Attitudes, Gender, and Political Trust As described above, while the Japanese have come to hold more progressive gender role views and have become more educated, Japanese women are still very disadvantaged in many areas. In spite of this trend, the LDP has hardly addressed policies on gender equality. For instance, the government set a rather conservative goal for maternity/ paternity leave usage. Stated target rates are 80% for maternity leaves and only 10% for paternity leaves. Moreover, the LDP eliminated “gender equality” from the party manifesto in 2005. Indeed, the party remains predominantly male, nominating only a fraction of women candidates in each election. Moreover, recent sexist comments by top leaders of the party send a message to the public that the very party that controls the government is insensitive to gender issues. Therefore, the fact that the government and its politicians appear insensitive to gender issues (and do not seem to be making much effort in improving gender relations) will not help them earn the trust of progressive individuals. We argue then, that there may be a wide gap between progressive individuals and the political leaders in their views on gender issues. In this circumstance, those who have progressive gender role attitudes may experience ‘relative deprivation,’ that is, the expectations of these progressive individuals are not met. This ‘relative deprivation’ ultimately leads to lower trust in members of the national parliament, that is, those who make policies. In other words, this distrust in political leaders may be attributed to ‘relative deprivation’ which progressive people experience, given the traditional nature of gender relations still present in Japan today. On the contrary, those with traditional gender role views are more likely to be satisfied with the political situation and more trusting of the incumbent politicians because the status quo in policy outcomes in gender relations is not much different from what they expect. Hence, these individuals are not as relatively deprived as their progressive counterparts. While it is more intuitive to expect progressive women to be relatively deprived, we expect that progressive men may also feel relative deprivation. Progressive men may also be dissatisfied without a gender-equal environment.

846

For instance, although the amendment to the original Child Care Leave Law of 1992 now allows employees (regardless of the size of the firm) to take child care leave, only a small fraction (.56%) of men utilized this benefit in 2004 (The Cabinet Office’s Gender Equality Committee on Declining Fertility and Gender-Equal Society 2005). Moreover, 48% of working men who did not take the leave reported the reason to be due to lack of support from their employers (The Japan Institute for Labor Policy and Training 2003). Therefore, the low rate of men taking parental leave reflects the traditional nature of the work environment in Japan. These facts suggest possible relative deprivation among both men and women because of the gap between their expectations and their surrounding reality. In sum, we expect that those who have progressive gender role attitudes are less likely to trust the members of the national parliament. Moreover women are the ones who have been directly impacted by gender equality policies and have been excluded from the circle of male dominated politics. Therefore, it is possible that women’s dissatisfaction has not only led to distrust in political leaders but has also extended to distrust in the system itself. Thus, our hypotheses are: H1: Those who have progressive gender role attitudes are less likely to trust the members of the national parliament than those who have traditional gender role attitudes. H2: Women who have progressive gender role attitudes are less likely to trust the government ministries and agencies than women who have traditional gender role attitudes. In testing these hypotheses, we examine the relationship between gender role attitudes, gender, and political trust. In addition, given their prior associations with political trust, we control for the following variables. First, generalized trust is controlled for to account for Erikson’s (1950) social–psychological theory. According to this theory, some people are generally more trusting due to their upbringing and such a trait simply influences their likelihood of trusting the government as well. Second, membership in civic and/or religious organizations is controlled to account for social capital theory. Social capital theory contends that engagement in civic activities generates a feeling of trust among participants, which translates into political trust (Brehm and Rahn 1997; Putnam 2000). Hence, based on social capital theory, those who belong to civic and religious organizations may be more likely to trust the government and politicians. Indeed, using time-series data, Keele (2007) has argued that declining social capital was partially responsible for the decrease in government trust despite improved government performance in the U.S. Moreover, those who are members of organizations may be more exposed

Sex Roles (2009) 60:843–858

to political cues, hence more likely to have an opinion. Third, we control for the respondents’ perceived financial situation and social class. In accordance with economic theory, scholars have found that one’s financial conditions influence their degree of confidence in the government (Citrin and Green 1986; Miller 1983; Weatherford 1984, 1987). Therefore, those who perceive their financial condition has worsened and those who perceive they belong to lower social class strata may be less likely to trust the government and government officials. Finally, we control for various demographic variables. Past literature on political trust in the U.S. has found no or very weak impact of individual characteristics on trust in the government and in politicians. However, Kabashima and his colleagues have found age, gender, education and place or residence to be related to trust in the government in Japan (Kabashima et al. 2000). They have found older people, men, those with more education and rural residents were more trusting of the government. Since the LDP controls the government, we should control for respondents’ party support or party identification. Unfortunately, these variables are not included in the dataset analyzed in our study. However, some of these individual characteristic variables have been found to covary with one’s party identification in Japan. For example, women are less likely to be supporters of the LDP party (Martin 2004) and supporters of the LDP are disproportionately drawn from rural areas (Flanagan et al. 1991). In sum, we control for measures of generalized trust, membership in civic and/or religious organizations, perceptions of personal financial situation, subjective social class, education, place of residence, and age.

Method Data We test our hypotheses using the 2003 Japanese General Social Survey (JGSS; Tanioka et al. 2003). The JGSS is based on a representative nationwide random sample of Japanese men and women 20 to 89 years of age who are living in Japan and registered to vote. The JGSS has been designed to gather political, sociological, economic, and demographic information. The JGSS has a face-to-face interview portion and a self-administered portion. It employs a split-half sample where half of the respondents are assigned to Form A and the other half are assigned to Form B. For the present study, we use responses from the Form A portion because it includes political trust variables and a battery of items that assess one’s gender role attitudes. The Form A portion of the JGSS yielded a response rate of 55% with completed surveys from 1,957 respondents (869 men and 1,088 women).

Sex Roles (2009) 60:843–858

Dependent Variables Trust in Members of the National Parliament/Trust in Government We use two measures of political trust as our dependent variables: trust in members of the national parliament and trust in government. Specific support is generally measured by inquiring of the respondents about their level of trust in current political leaders as members of parliaments. Diffused support is measured by asking respondents about their level of trust in political institutions such as the government (Dalton 2006). We follow this common practice. In the 2003 JGSS, respondents were given a list of items and were asked how much they trust them. We use two items among the list: “Diet members” and “ministries and government agencies.” Diet members refer to the members of the national parliament in Japan. Valid responses were “very much,” “some,” “not very much,” and “don’t know.” The “very much” category yielded only a very small percentage of the sample on both political trust variables (1.54% for trust in members of the national parliament and 4.62% for trust in government). This may be due to respondents avoiding responses at the extreme end of the scale. Therefore, we collapsed “very much” and “some” into one category representing those who trust. The two political trust variables resulted in having three categories: 0=don’t know, 1=some or very much, and 2= not very much. Conventional studies treat “don’t know” (DK) responses as missing values and exclude them from analyses. We opt to treat DK as a valid response because quite a large percentage of respondents reported DK on the political trust questions (12.06% of men and 22.70% of women on the trust in members of the national parliament question, and 16.42% of men and 31.01% of women on the trust in government question). DK answers may have qualitatively different meanings. DK responses to opinion and attitude questions can mean that they have no idea, feel unsure, or are undecided. To prevent clueless respondents from randomly choosing one of the substantive options, it is recommended that surveys provide explicit “don’t know” and “no opinion” options and to avoid persistent probing (Sanchez and Morchio 1992). In the JGSS, the political trust questions explicitly offered the “don’t know” option and were not subject to probing because they were part of the self-administered part of the survey. For our purposes, distinguishing those with and without a political opinion is valuable in determining what influences people to have an opinion in the first place. We argue that those who report DK represent those who do not have sufficient knowledge and/or interest about national politics to form an opinion about political trust. To examine

847

this idea, we looked at the distribution of the responses to the question, “regardless of whether it is the time of an election or not, some people show continuous interests in politics and others don’t. How regularly do you pay attention to the political situation?” This question was asked to the other half of the 2003 JGSS sample (those who were assigned to Form B). 36.3% of men and 50.8% of women responded that they “rarely” or “hardly ever” pay attention to political situations. Thus, high occurrences of DK responses to the political trust questions are reasonable, given such a widespread lack of political interest among the Japanese. Moreover, that a greater percentage of women exhibit less interest in politics than men is consistent with the fact that more women chose DK than their male counterparts. In this paper, those who are in the DK category are also referred to as those without an opinion. Independent Variables Gender Role Attitudes Respondents’ gender role attitudes are measured using responses to a series of items included in the 2003 JGSS that assess one’ views of gender roles. The survey includes a wide range of items assessing individual attitudes toward divorce and marriage, and attitudes toward having children. However, we only use the following six statements that directly measure one’s gender role views. & & & & & &

If a husband has sufficient income, it is better for his wife not to have a job. Without a doubt, a woman’s happiness lies in a marriage. Men should cook and look after themselves. (reverse coded) A husband’s job is to earn money; a wife’s job is to look after the home and family A preschool child is likely to suffer if his or her mother works. It is more important for a wife to help her husband’s career than to have one herself.

Responses were coded as follows: ranged from 1=agree, 2=somewhat agree, 3=somewhat disagree, and 4=disagree. After reverse coding the third item to ensure that higher responses indicate more progressive views, we summed responses of all items and created a gender role attitude measure. This measure ranged from 6 to 24 where lower values indicate more traditional values and higher values indicate more progressive gender role attitudes. Cronbach’s alpha for this summated scale was .78 for the overall sample. When computed separately for men and women, they were .77 and .78, respectively.

848

Gender Since we expect that the effect of gender role attitudes on political government differs between women and men, we will include the respondents’ gender (coded as 1=women and 0=men). We further explore our association of interest by running our models separately for men and women.

Sex Roles (2009) 60:843–858

a series of dummy variables for last school the respondents attended. Dummy variables were each created for junior high school, high school, and post-secondary school. Respondent’s place of residence is also dummy coded 1 where respondents reside in one of the 13 major cities or other cities and 0 represents residence in a town or a village. Analysis Plan

Control Variables Generalized Trust Generalized trust is measured by the question, “generally speaking, would you say that most people can be trusted?” While original responses were “yes,” “no,” and “depends,” they were recoded to create a dummy variable where 1 represents those who trust and 0 represents those who do not necessarily always trust. Organizational membership Membership in civic and religious organizations is measured with a dummy variable where 1 means that the respondent is a member of at least one of the following three organizations: consumers’cooperative society, hobby groups and clubs (choir, photography, hiking, etc.), and religious groups. Respondents who are not a member of any of these groups are given a 0 and are the reference group. Personal Financial Situation Perception of personal financial situation is measured by the question, “During the last few years, has your financial situation been getting better, worse, or has it stayed the same?” Original responses were “getting better,” “getting worse,” and “stayed the same.” These responses were recoded into a dummy variable where 1 was assigned to those who reported their financial situation as getting worse, and 0 was assigned to those who indicated either “getting better” or “stayed the same.”

All results are based on a sample of 1,807 JGSS respondents (804 men and 1,003 women) for whom we have data on all of the variables included in the analyses. Of the 1,957 interviews completed for Part A of the JGSS, 150 cases were excluded due to missing values. Our analysis has four parts. First, we show descriptive statistics of all variables and present differences in distributions by gender. We also present bivariate analysis showing differences in political trust by gender according to our independent variables. Then, we conduct multinomial logistic regression analyses to estimate the effects of independent variables on the odds of trusting, distrusting, or not having an opinion about the members of the national parliament and the government. In these analyses, we first run an initial model (Model 1) that assesses the effects of independent variables derived from past research on political trust. In the proceeding model (Model 2), we add gender role attitudes to the equation to assess our contribution. Finally, we examine the association between political trust and all independent variables for men and women separately. For each equation, variance inflation factors (VIFs) were used to detect problems of multicollinearity. As suggested by low VIFs, multicollinearity problems did not exist (all VIFs were under 2.5). Since the 2003 JGSS employed two-stage stratified random sampling, all multivariate analyses are conducted using appropriate survey weights and survey estimation procedures in Stata 10.0.

Results

Subjective Social Class

Descriptive and Bivariate Statistics

Subjective social class was measured by the question “If we were to divide the contemporary Japanese society into the following five strata, which would you say you belong to?” Responses ranged from 1 (upper) to 5 (lower).

Table 1 shows descriptive statistics of all variables included in our study by gender. Overall, a greater percentage of respondents (53.02%) do not trust the members of national parliament whereas only 29.00% trust and only 17.99% had no opinion on them. Looking at trust in the government, we see that more people trust the government (44.27%) than distrust the government (31.21%) or have no opinion (24.52%). Thus members of the national parliament are generally perceived more unfavorably than the government.

Demographic Variables (Age, Education, and Place of Residence) Age is measured as a continuous variable representing respondent’s age in years. To control for education, we created

Sex Roles (2009) 60:843–858

849

Table 1 Descriptive statistics of dependent and independent variables by gender. Total

Women

Men

Percentages Dependent variables Trust in diet membersa 0 = Trust 1 = Distrust 2 = Do not know Trust in governmenta 0 = Trust 1 = Distrust 2 = Do not know Independent variables Gender role attitudesa (higher values indicate more progressive attitudes) Generalized trust 1 = Trust 0 = Can’t trust/depends Membership 1 = Membershipa 0 = Not member Financial situation 1 = Getting worse 0 = Getting better/same Subjective social class 1 = Upper 2 = Upper middle 3 = Middle middle 4 = Lower middle 5 = Lower Last School attended 1 = Junior high, 0 = other 1 = High school, 0 = othera 1 = Post-secondary, 0 = other Place of residence 1 = City 0 = Town/village Age n

29.00 53.02 17.99

28.91 48.35 22.73

29.10 58.83 12.06

44.27 31.21 24.52

42.97 26.02 31.01

45.90 37.69 16.42

21.42 78.58

20.24 79.76

22.89 77.11

38.90 61.10

44.77 55.23

31.59 68.41

51.80 48.20

50.15 49.85

53.86 46.14

.33 9.46 49.97 53.04 6.97

.20 8.97 53.04 31.41 6.38

.50 10.07 46.14 35.57 7.71

18.15 52.30 29.55

17.05 54.44 28.51

19.53 49.63 30.85

76.43 23.57

78.07 21.93

74.38 25.62

Total Range

Mean

6–24

15.44

1,003

SD

3.80

Mean

15.94

Men SD

3.82

Mean

14.82

0–1

.21

.20

.23

0–1

.39

.45

.32

0–1

.52

.50

.54

1–5

3.37

0–1 0–1 0–1 0–1

.18 .52 .30 .76

20–89 1,807

Women

52.60

.76

3.35

.74

.17 .54 .29 .78

16.77

52.44

3.40

SD

3.68

.79

.20 .50 .31 .74

16.59

52.79

17.00

804

Percentages, range, mean, and SD indicated only if applicable a Statistically significant differences by gender

When we focus on gender differences in political trust, we see that men are more distrusting of both members of the national parliament and the government. The percentage point differences between men and women (M-W) who distrust are 10.48 percentage points for members of the national parliament and 11.67 percentage points for the government. There is also a noticeable gender difference in the percentages of those who have no opinion. More women do not have any opinion compared to men (22.73% vs. 12.06% for trust in members of the national parliament and 31.01% vs. 16.42% for trust in government). When we exclude DK respondents, the percentage point differences between men and women (M-W) who distrust decreases to 5.29% points and to 7.37% points for

members of the national parliament and the government, respectively. In terms of gender role attitudes, we see that women have more progressive gender role attitudes than men as indicated by their significantly higher scores on the gender role attitude measure. Mean scores on the gender role attitude scale are significantly different for men and women. On a scale that ranges from 6 to 24 where higher values indicate more progressive gender role attitudes, means were 14.82 and 15.94 for men and women, respectively. Table 2 and 3 show means and percentages of the respondents who trust, distrust, or have no opinion on the members of the national parliament and the government according to independent variables. Examining differences in

a

44.19 55.42 53.62 52.06 56.94 48.79 33.33 48.54 48.62 59.57 60.32 43.90 51.53 61.24 55.03 46.48

27.08 32.01

26.28 31.92

50.00 35.09 32.78 23.63 18.25

34.76 28.78 25.84

28.46 30.75

16.07 (3.84) 49.80 (16.48)

40.31 25.92

14.35 (3.59) 57.58 (16.18)

Higher values indicate more progressive attitudes

Percentages Generalized trust Trust Can’t trust/depends Membership Member Not member Financial situation Getting worse Getting better/same Subjective social class Upper Upper middle Middle middle Lower middle Lower Last school attended Junior high High school Post secondary Place of residence City Town/village

Means (SD) Gender role attitudesa Age

16.51 22.77

21.34 19.68 12.92

16.67 16.37 18.60 16.81 21.43

16.77 19.29

15.93 19.29

15.50 18.66

15.36 (3.60) 54.07 (17.26)

29.37 27.27

36.26 27.84 26.57

100.00 33.33 33.27 22.54 15.63

26.64 31.20

32.70 24.92

40.39 26.00

14.70 (3.84) 57.49 (15.99)

Trust

DK

Trust

Distrust

Women

Total

49.81 43.18

38.01 47.62 55.94

.00 44.44 44.17 54.29 60.94

53.08 43.60

46.62 48.65

41.38 50.13

16.81 (3.71) 49.25 (16.39)

Distrust

20.82 29.55

25.73 24.54 17.48

.00 22.22 22.56 23.17 23.44

20.28 25.20

20.68 26.43

18.23 23.88

15.67 (3.57) 54.39 (16.74)

DK

44.65 49.51

33.12 30.08 25.00

25.00 37.04 32.08 24.83 20.97

25.87 32.92

29.67 28.99

40.22 25.81

13.90 (3.20) 57.69 (16.45)

Trust

Men

Table 2 Means and percentage distributions by trust in members of the national parliament and by gender according to independent variables.

38.63 34.95

50.32 56.89 67.34

50.00 53.09 54.99 65.38 59.68

61.43 54.95

61.17 56.78

47.28 62.26

15.31 (3.83) 50.37 (16.56)

Distrust

16.72 15.53

16.56 13.03 7.66

25.00 9.88 12.94 9.79 19.35

12.70 11.32

9.16 14.24

12.50 11.94

14.65 (3.20) 53.32 (18.47)

DK

850 Sex Roles (2009) 60:843–858

a

25.06 32.89 31.44 31.07 35.58 26.52 .00 29.82 26.25 37.10 42.06 21.95 31.01 37.27 32.22 27.93

47.94 41.94

40.17 48.68

50.00 56.73 49.28 35.77 31.75

41.16 44.87 45.13

43.74 46.01

16.08 (3.83) 48.66 (16.18)

57.11 40.77

14.82 (3.76) 55.45 (16.18)

Higher values indicate more progressive attitudes.

Percentages Generalized trust Trust Can’t trust/depends Membership Member Not member Financial situation Getting worse Getting better/same Subjective social class Upper Upper middle Middle middle Lower middle Lower Last school attended Junior high High school Post secondary Place of residence City Town/village

Means (SD) Gender role attitudesa Age

24.04 26.06

36.89 24.13 17.60

50.00 13.45 24.47 27.12 26.19

24.25 24.80

20.63 26.99

17.83 26.34

15.58 (3.69) 52.44 (18.52)

43.04 42.73

38.10 43.28 25.38

100.00 52.13 47.55 33.53 30.14

38.97 47.00

46.33 40.25

56.16 39.63

15.31 (3.94) 54.33 (16.34)

Trust

DK

Trust

Distrust

Women

All

27.33 21.36

16.93 26.21 67.42

.00 24.47 22.38 29.64 35.62

30.82 21.20

28.06 24.37

22.66 26.88

16.88 (3.67) 49.30 (14.76)

Distrust

Table 3 Means and percentage distributions by trust in government and by gender according to independent variables.

29.63 35.91

44.97 30.52 7.20

.00 23.40 30.07 36.83 34.25

30.22 31.80

25.61 35.38

21.18 33.50

16.03 (3.61) 52.46 (17.98)

DK

44.65 49.51

44.59 46.37 45.97

25.00 61.73 50.94 38.11 32.26

41.57 50.94

51.28 43.22

58.15 42.26

14.46 (3.49) 56.76 (15.92)

Trust

Men

38.63 34.95

26.75 37.09 45.56

.00 34.57 31.27 45.80 45.16

41.11 33.69

37.36 37.22

27.72 40.65

15.39 (3.84) 48.12 (15.75)

Distrust

16.72 15.53

28.66 16.54 8.47

75.00 3.70 17.79 16.08 22.58

17.32 15.36

11.36 19.55

14.13 17.10

15.55 (3.68) 52.41 (19.82)

DK

Sex Roles (2009) 60:843–858 851

a

^

1.16 1.20

1.27 .98 *** 8.66 (20) 1,807

1.36***

1.21 1.34

1.29^ .98 *** 8.34 (18) 1,807

1.13

1.14 1.87*

1.64 ** .99 8.66 (20) 1,807

1.35***

.96* 1.23^

.98 1.28

1.39* 1.27

.73* .51***

.80^ .55***

1.10***

Model 2

.39*** .85

1.04^

Model 1

p≤.1, two-tailed test; *p≤.05, two-tailed test; **p≤.01, two-tailed test; ***p≤.001, two-tailed test; Higher values indicate more progressive attitudes

Gender role .94** attitudesa Gender .50*** .53*** .41*** Generalized 1.59* 1.65** .87 trust 1.41* Membership 1.43* 1.45^ Financial 1.00 1.03 1.29^ situation Subjective .84 .84 1.14 social class Last school attended Junior high school (reference) High school .97 .99 1.17 1.96 ** Post1.46 1.56^ secondary school Place of residence Town/village (reference) City 1.27 1.29 1.64 ** Age 1.02 ** 1.01 * .99^ F (df) 8.34 8.66 8.34 (18) (20) (18) N 1,807 1,807 1,807

Model 2

1.09 1.02 *** 12.55 (18) 1,807

1.94*** 3.08***

.79*

1.47** .94

.43*** 1.87***

Model 1

1.10 1.02 *** 11.76 (20) 1,807

1.97*** 3.12***

.79*

1.49** .95

.45*** 1.90***

.97^

Model 2

Model 1

Model 1

Model 2

Trust vs. DK

Distrust vs. DK

Trust vs. DK

Distrust vs

Trust in government

Trust in member of the National Parliament

Table 4 Relative risk ratios from multinomial logistic regression analyses predicting political trust (full sample).

1.42 * 1.00 12.55 (18) 1,807

2.28*** 3.46***

1.15

1.58** 1.35*

.29*** 1.42*

Model 1

1.42 1.00 11.76 (20) 1,807

2.23*** 3.33***

1.15

1.57** 1.34

.28*** 1.05

1.02

Model 2

Distrust vs. DK

1.30^ .98 *** 12.55 (18) 1,807

1.18 1.12

1.45 ***

1.07 1.44**

.67*** .57***

Model 1

Distrust vs

1.29^ .98 *** 11.76 (20) 1,807

1.13 1.04

1.63 ***

1.06 .55**

.63*** .55***

1.06**

Model 2

852 Sex Roles (2009) 60:843–858

Sex Roles (2009) 60:843–858

gender role attitudes among the three political trust categories, those who do not trust their political leaders or the government on average have the most progressive gender role attitudes as compared to those who trust or who do not have an opinion. This pattern is evident among both women and men. Multivariate Results Table 4 presents results from our multinomial logistic regression models which examine the impact of the independent variables on trust in their political leaders and trust in the government for a combined sample of men and women. We report relative risk ratios which indicate the odds that members with a given attribute report a given response to the political trust questions relative to a reference group controlling for all other predictors. With three possible outcomes (trust, distrust, and DK), we first contrast those who have an opinion (either trust or distrust) to those who do not have an opinion to determine factors associated with having an opinion. Then, we contrast those who distrust to those who trust to examine factors that contribute to an unfavorable evaluation. Model 1 presents findings from our baseline model where we estimate the effects of variables derived from past theories and research on the odds of trusting, distrusting, and having no opinion of political trust. As we have seen in our bivariate results, women are significantly less likely to have an opinion on political trust than men. We also see that among those who have an opinion, women in general are more likely to trust the government compared to men. As for membership in organizations, those who are members have greater odds of trusting or distrusting as opposed to not having an opinion. Through membership, perhaps individuals are more exposed to political cues, hence can form an opinion about the political incumbents and the government. This could also be because individuals who are politically active join such organizations, thus are more likely to hold an opinion. As we hypothesized, we also find that having generalized trust is associated with higher odds of trusting political leaders and the government, as opposed to distrusting or having no opinion. As expected, those who perceive their financial situation has worsened during the last few years are more likely to distrust members of the national parliament and the government. Having a lower subjective social class is also associated with not having faith in their political leaders and the government. In terms of demographic variables, estimated effects of the education variables suggest that greater education increases the odds of trusting or distrusting the members of the national parliament and the government than reporting DK. This finding is consistent with past research on education and political knowledge (Delli Carpini and Keeter 1993, 1996; Mondak 1999) in that those who lack

853

political knowledge are less likely to take political cues from which they can evaluate the government. However, we see that education is not associated with whether one perceives their political leaders and the government favorably or unfavorably. Increases in age are associated with increases in the odds of trusting their political leaders and the government. Older people have experienced the economic reconstruction after WWII under the LDP’s leadership. Thus, those who witnessed this so-called “economic miracle” may still positively evaluate the LDP’s leadership even though the economy has been stagnating for over a decade and numerous political scandals have surfaced since the 1990s. Finally, city dwellers have a higher likelihood of distrusting members of the national parliament and the government as compared to town/village residents. This is possibly because city dwellers are more likely to be supporters of the opposition parties (Richardson 1997) and are more likely to be dissatisfied with the current administration and the government led by the LDP. Model 2 presents our added contribution of one’s gender role attitudes. Most significant associations between the political trust variables and independent variables from the baseline model remained significant after introducing the gender role attitude measure. Comparing those who do not trust their political leaders and the government to those who trust, respondents who scored higher on the gender role attitude measure have greater odds of not trusting. In other words, Japanese who have progressive gender role attitudes tend to evaluate their political leaders and the government negatively. We believe this highlighted relationship deserves more investigation. In the next series of analyses, we examine men and women separately to further explore the association between gender role attitudes and political distrust. Gender Differences in Political Trust Tables 5 and 6 address whether the relationship between one’s gender role attitudes and political trust is the same for men and women. Once men and women are analyzed separately, we see a slightly different pattern between the genders. Results from our analysis of both men and women suggest that progressive gender role attitudes increase the odds of not trusting members of the national parliament as opposed to trusting them, lending support to H1. In addition, progressive gender role attitudes are associated with a higher likelihood of evaluating the government negatively than positively among women. This supports H2. Thus, women who have progressive gender role attitudes are more likely to evaluate their political leaders and the government negatively. A parallel finding to this is that women with traditional gender role views tend to trust

854

Sex Roles (2009) 60:843–858

Table 5 Relative risk ratios from multinomial logistic regression analyses predicting political trust (women only). Trust in members of the National Parliament

Trust in government

Trust vs. DK

Distrust vs. Trust

Trust vs. DK

Distrust vs. DK

Distrust vs. Trust

1.13*** .58** .73^ 1.28 1.50***

.95* 2.16*** 1.43* 1.01 .75**

1.04 1.32 1.56* 1.59* 1.11

1.09*** .61* 1.09 1.57** 1.48**

1.28 1.18

1.72* 2.15**

2.05** 2.34**

1.19 1.09

1.10 .98*** 6.21 (18) 1,003

1.39 1.01* 5.28 (18) 1,003

1.89** 1.00 5.28 (18) 1,003

1.36 .99* 5.28 (18) 1,003

Distrust vs. DK

Gender role attitudesa .92** 1.04 Generalized trust 1.80* 1.04 Membership 1.70** 1.24 Financial situation 1.21 1.54* 1.16 Subjective social class .77^ Last school attended Junior high school (reference) High school .87 1.12 Post-secondary school 1.29 1.53 Place of residence town/village (reference) City 1.62* 1.79** Age 1.01 .99^ F (df) 6.21(18) 6.21 (18) N 1,003 1,003 ^ a

p≤.1, two-tailed test; *p≤.05, two-tailed test; **p≤.01, two-tailed test; ***p≤.001, two-tailed test Higher values indicate more progressive attitudes

their political leaders and the government. For traditional women, the status quo in gender relations is perhaps not much different from what they expect, thus they are more likely to be satisfied with the political situation. For men, as expected, exhibiting progressive gender role attitudes increases the odds of distrusting their political leaders as opposed to trusting them. This provides additional support for our first hypothesis. However, unlike for women, this distrust does not extend to distrust in the system itself. This supports our argument that because gender issues are more salient among women, distrust in members of the national parliament extends to distrust in the government only among women.

For both men and women, generalized trust, subjective social class, and age are associated with the outcome variables in the expected way. For both men and women, greater education increases the odds of having an opinion about their political leaders and the government. However, levels of education are not related to whether one trusts or distrusts members of the national parliament. This is somewhat puzzling. We need further research to investigate why education does not help the Japanese evaluate their political leaders. The impact of one’s perception of changes in their financial situation on political trust exists only among women. We find that women who perceive that their

Table 6 Relative risk ratios from multinomial logistic regression analyses predicting political trust (men only). Trust in members of the National Parliament

Trust in Government

Trust vs. DK

Distrust vs. Trust

Trust vs. DK

Distrust vs. DK

Distrust vs. Trust

1.08** .48*** 1.26 1.19 1.22^

.99 1.60^ 1.58^ .85 .86

1.02 .81 1.60^ 1.10 1.21

1.03 .51** 1.01 1.29 1.41**

1.05 1.15

2.33** 5.67***

2.45** 5.56***

1.05 .96

1.47^ .98** 2.99 (18) 804

.70 1.03*** 5.57 (18) 804

.90 1.00 5.57 (18) 804

1.29 .97*** 5.57 (18) 804

Distrust vs. DK

.97 1.04 Gender role attitudesa Generalized trust 1.34 .65 Membership 1.23 1.55 Financial situation .78 .93 Subjective social class .91 1.11 Last school attended Junior high school (reference) High school 1.11 1.17 2.38* Post-secondary school 2.08^ Place of residence town/village (reference) City .90 1.32 Age 1.02 1.00 F (df) 2.99 (18) 2.99 (18) N 804 804 ^ a

p≤.1, two-tailed test; *p≤.05, two-tailed test; **p≤.01, two-tailed test; ***p≤.001, two-tailed test Higher values indicate more progressive attitudes

Sex Roles (2009) 60:843–858

economic situation has worsened (compared to women who report that the situation has not changed or gotten better) are more likely to evaluate the government negatively. These women are also at greater odds of distrusting the members of the national parliament than not having an opinion about them. That the impact of this variable is significant only among women is possibly because Japanese women are often responsible for their household finances (Iwao 1993; Robins-Mowry 1983), hence sensitive to their financial situation. Indeed, it has been reported that Japanese women are more concerned about tax issues than men (Patterson and Nishikawa 2002).

Discussion We found that gender role attitudes have independent effects on political trust among the Japanese. Both men and women who have progressive gender role attitudes are less likely to trust the members of the national parliament, controlling for other factors. Moreover, this distrust extends to the government itself among women. This implies that gender equality issues are a potential political cleavage among the Japanese electorate. In spite of these findings, gender equality issues are almost non-existent in Japanese political dialogues. Indeed, “gender equality” was dropped from the LDP’s manifesto in 2005. We suggest that the absence of this issue on the agenda is a source of women’s political distrust. Thus, addressing gender issues in Japan can be used to incorporate Japanese women into the political arena. Currently Japanese women are politically under-mobilized. This is reflected in the fact that only a fraction of women are among members of the national parliament and fewer women show interest in politics than men. Since the Japanese are becoming progressive, the finding that progressive individuals are less likely to have political trust has a significant implication. The number of Japanese who are relatively deprived may increase with an increase in education. Relative deprivation may also increase in degree since the Japanese are becoming ever more progressive. Hence, without implementation of policies to advocate gender-equality in society, political distrust among the Japanese may persist in the future.

855

It is ideal to estimate relative deprivation by examining one’s evaluation of government performances on gender issues and their gender role attitude. However, the 2003 JGSS does not ask respondents to evaluate the government’s (or government officials’) performances on gender issues. Therefore, we used gender role attitude as a proxy of the respondents’ evaluation of the government (government officials’) performance on gender issues. Given that the Japanese society is still traditional in terms of gender relations, we assumed that those with progressive gender role attitudes evaluate the government differently than those with traditional values on gender issues. We recognize that this is one of the limitations of our study. In addition, ideally, respondent’s party identification should be controlled for. Party identification was unfortunately not included in Form A of the JGSS. However, we believe that we successfully controlled for party identification by including two variables that are known to be associated with support for the LDP (i.e. gender and place of residence). Our study is limited to Japan. Women in other countries are also becoming increasingly educated and progressive but some countries accommodate their emerging needs to improve gender relations better than others. Future research should explore the impact of gender role attitudes and political trust comparatively using data such as the World Values Survey (WVS). Particularly, it will be interesting to compare the impact of one’s gender role attitudes on political trust between countries where women enjoy relatively equal gender relations (e.g. Scandinavian countries) and countries where women are still relatively disadvantaged in social, economic and political spheres (e.g. Japan, Korea, and Italy). Acknowledgements The authors wish to thank Edward Crenshaw for his helpful comments on an earlier draft of this paper. We also thank Charles Smith for proofreading and providing insightful comments. The Japanese General Social Surveys (JGSS) analyzed in this paper are designed and carried out at the Institute of Regional Studies at Osaka University of Commerce in collaboration with the Institute of Social Science at the University of Tokyo under the direction of Ichiro TANIOKA, Michio NITTA, Hiroki SATO and Noriko IWAI with Project Manager, Minae OSAWA. The project is financially assisted by Gakujutsu Frontier Grant from the Japanese Ministry of Education, Culture, Sports, Science and Technology for 1999-2003 academic years, and the datasets are compiled with cooperation from the SSJ Data Archive, Information Center for Social Science Research on Japan, Institute of Social Science, the University of Tokyo.

856

Appendix 1: Original Survey Questions From the 2003 JGSS Used in the Study

Sex Roles (2009) 60:843–858

Q9. If we were to divide the contemporary Japanese society into the following fives strata, which would you say you belong to? 1 Upper 2 Upper middle 3 Middle middle 4 Lower middle 5 Lower Q13. Generally speaking, would you say that most people can be trusted? 1 Yes 2 No 3 Depends Q15. How well do you trust... D. Ministries and government agencies? K. Diet members? 1 Very much 2 Some 3 Not very much 4 Don’t know Q19. Do you agree or disagree with the following statements? B. If a husband has sufficient income, it is better for his wife not to have a job. C. Without a doubt, a woman’s happiness lies in a marriage. D. Men should cook and look after themselves. E. A husband’s job is to earn money; a wife’s job is to look after the home and family. G. A preschool child is likely to suffer if his or her mother works. I. It is more important for a wife to help her husband’s career than to have one herself. 1 Agree 2 Somewhat agree 3 Somewhat disagree 4 Disagree Q20. Are you a member of the following organizations? A. Cooperatives society (co-op) B. Religious groups C. Hobby groups and clubs (choir, photo-taking, hiking, etc) 1 Yes 2 No Q25. What was the last school you attended (or are attending now)? (Before World War II)

English Translation Q3. During the last few years, has your financial situation been getting better, worse, or has it stayed the same? 1 Getting better 2 Getting worse 3 Stayed the same

(a) Ordinary elementary school in the old system (including national elementary school) (b) Higher elementary school in the old system (c) Junior high school/Girls’ high school in the old system (d) Vocational school in the old system (e) Normal school in the old system

Sex Roles (2009) 60:843–858

(f) Higher school or vocational school in the old system/higher normal school (g) University/Graduate school in the old system (After World War II) (h) Junior high school (i) High school (j) 2-year college/College of technology (k) University (l) Graduate school

References Anderson, C. J., & Guillory, C. (1997). Political institutions and satisfaction with democracy: a cross-national analysis of consensus and majoritarian systems. American Political Science Review, 91, 66–81. Beech, H. (2005, August 29). The wasted asset: Japanese women are smart and entrepreneurial, so why is so little effort made to harness their talents? Time International (Asia Edition), 166(9), 34 Retrieved March 19, 2008, from http://find.galegroup.com/itx/ start.do?prodId=AONE. Brehm, J., & Rahn, W. M. (1997). Individual-level evidence for the causes and consequence of social capital. American Journal of Political Science, 41, 999–1023. Cabinet Office of Japan. (2002). Danjo kyoudou sanakaku ni kansuru yoron chousa. [Opinion survey on gender equality]. Retrieved March 19, 2008, from http://www8.cao.go.jp/survey/h14/h14-danjo/index. html. Cabinet Office of Japan. (2006). Danjo kyoudou sankaku hakusho. [White paper on gender equality]. Retrieved March 21, 2008, from http://www.gender.go.jp/whitepaper/h18/web/danjyo/html/ zuhyo/fig01_08_01.html. Cabinet Office’s Gender Equality Committee on Declining Fertility and Gender-Equal Society. (2005). Shoushika to danjo kyoudou sankaku ni kansuru shakai kankyou no kokusai hikaku houkokusho. [An international comparison on declining birth rate, gender equality and social environment]. Retrieved October 28, 2008, from http://www.gender.go.jp/danjo-kaigi/ syosika/houkoku/index-kokusai.html. Canadian Association of University Teachers. (2004). CAUT Almanac of post-secondary education in Canada. Retrieved August 23, 2007, from http://hdl.handle.net/1873/3614. Citrin, J. (1974). Comment. The political relevance of trust in government. American Political Science Review, 68, 973–988. Citrin, J., & Green, D. P. (1986). Presidential leadership and the resurgence of trust in government. British Journal of Political Science, 16, 431–453. Davis, S. N., & Greenstein, T. N. (2004). Cross-national variations in the division of household labor. Journal of Marriage and the Family, 66, 1260–1272. Dalton, R. (2006). Citizen politics: public opinion and political parties in advanced industrial democracies.(4th ed.). Washington DC: CQ Press. Deitch, C. (1988). Sex differences in support for government spending. In C. W. Mueller (Ed.), The Politics of the Gender Gap (pp. 192–216). Newbury Park, CA: Sage.

857 Delli Carpini, M. X., & Keeter, S. (1993). Measuring political knowledge: Putting first things first. American Journal of Political Science, 37, 1179–1206. Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it matters. New Haven, CT: Yale University Press. Easton, D. (1965). A system analysis of political life. New York, NY: Wiley. Easton, D. (1975). A reassessment of the concept of political support. British Journal of Political Science, 5, 435–457. Endersby, J. W., & Krieckhaus, J. T. (2008). Turnout around the globe: The influence of electoral institutions on national voter participation, 1972–2000. Electoral Studies, 27, 601–610. Erikson, E. H. (1950). Childhood and society. New York, NY: Norton. Flanagan, S. C., Kohei, S., Miyake, I., Richardson, B. M., & Watanuki, J. (1991). The Japanese voter. New Haven, CT: Yale University Press. Gurr, T. (1968). A causal model of civil strife: a comparative analysis using new indices. The American Political Science Review, 62, 1104–1124. Hibbing, J., & Theiss-Morse, E. (1995). Congress as public enemy: public attitudes toward American political institutions. New York, NY: Cambridge University Press. Inter-Parliamentary Union. (2007). Women in national parliaments. Retrieved August 23, 2007, from http://www.ipu.org/wmn-e/ classif.htm. Iwao, S. (1993). The Japanese women: Traditional image and changing reality. New York, NY: Free Press. Japan minister rapped on women (2007, January 29). London, UK: BBC News, Retrieved March 19, 2008, from http://news.bbc.co. uk/2/hi/asia-pacific/6309405.stm. Japan Institute for Labor Policy and Training.(2003). Ikuji ya kaigo to shigoto no ryouritsu ni kansuru chousa. [A study on child care and employment]. Retrieved October 28, 2008, from http://www. jil.go.jp/kokunai/statistics/doko/h1507/subindex.html. Kabashima, I., Marshall, J., Uekami, T., & Hyun, D. (2000). Casual cynics or disillusioned democrats? Political alienation in Japan. Political Psychology, 21, 779–804. Keele, L. (2007). Social capital and the dynamics of trust in government. American Journal of Political Science, 5, 241–254. LaFont, S. (2001). One step forward, two steps back: women in the post-communist states. Communist and Post-Communist Studies, 34, 203–220. Martin, S. L. (2004). Alienated, independent and female: Lessons from the Japanese electorate. Social Science Japan Journal, 7, 1–19. Miller, A. H. (1974). Political issues and trust in government: 1964– 1970. The American Political Science Review, 68, 951–972. Miller, A. H. (1983). Is confidence rebounding? Public Opinion, 6, 16–20. Miller, W. E., Miller, A. H., & Schneider, E. J. (1980). American national election studies data sourcebook 1852–1978. Cambridge, MA: Harvard University Press. Mondak, J. J. (1999). Reconsidering the measurement of political knowledge. Political Analysis, 8, 57–82. National Institute of Population and Social Security Research (2007). Report on the thirteenth Japanese national fertility survey, 2005, Volume II: Attitudes toward marriage and family among Japanese singles. Tokyo, Japan: National Institute of Population and Social Security Research (In Japanese). OECD (2004). Education at a glance: OECD indicators. Paris, France: Organisation for Economic Co-operation and Development. Page, B. I., & Shapiro, R. Y. (1992). The rational public: Fifty years of trends in Americans’ policy preferences. Chicago, IL: University of Chicago Press.

858 Patterson, D., & Nishikawa, M. (2002). Political interest or interest in politics? Gender and party support in postwar Japan. Women & Politics, 24, 1–34. Pharr, S. J., & Putnam, R. D. (2000a). Disaffected democracies: what’s troubling the trilateral countries? Princeton, NJ: Princeton University Press. Pharr, S. J., & Putnam, R. D. (2000b). Introduction: What’s troubling trilateral democracies? In S. J. Pharr, & R. D. Putnam (Eds.), Disaffected democracies: what’s troubling the trilateral countries? (pp. 3–30). Princeton, NJ: Princeton University Press. Powell Jr., G. B. (1986). American voter turnout in comparative perspective. The American Political Science Review, 80, 17–43. Putnam, R. P. (2000). Bowling alone. New York, NY: Simon & Schuster. Richardson, B. (1997). Japanese democracy: Power, coordination, and performance. New Haven, CT: Yale University Press. Robins-Mowry, D. (1983). The hidden sun: Women of modern Japan. Boulder, CO: Westview Press. Sanchez, M. E., & Morchio, G. (1992). Probing “Don’t know” answers: Effects on survey estimates and variable relationships. Public Opinion Quarterly, 56, 454–474. Tanioka, I., Iwai, N., Nitta, M., & Sato, H.(2003). JAPANESE GENERAL SOCIAL SURVEY (JGSS) [Computer file].

Sex Roles (2009) 60:843–858 ICPSR04242-v1. Osaka, Japan: Osaka University of Commerce, Office of Japanese General Social Surveys [producer], 2005. Ann Arbor MI: Inter-university Consortium for Political and Social Research [distributor], 2005–09–30. Treas, J., & Widmer, E. D. (2000). Married women’s employment over the life course: Attitudes in cross-national perspective. Social Forces, 78, 1409–1436. United Nations Statistics Division.(2007). Statistics and indicators on women and men. Retrieved March 12, 2008, from http://unstats. un.org/unsd/demographic/products/indwm/tab5d.htm. U.S. Department of Education National Center for Education Statistics. (2003). Integrated postsecondary education data system (IPEDS), Winter 2001–02. Retrieved March, 12, 2008, http://nces.ed.gov/programs/digest/d03/tables/dt231.asp. Walker, I., & Smith, H. J. (2002). Relative deprivation: specification, development, and integration. Cambridge, New York: Cambridge University Press. Weatherford, S. M. (1984). Economic ‘stagflation’ and public support for the political system. British Journal of Political Science, 14, 187–205. Weatherford, S. M. (1987). How does government performance influence political support? Political Behavior, 9, 5–28. Wirls, D. (1986). Reinterpreting the gender gap. Public Opinion Quarterly, 50, 316–330.