PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

7 ro‘ffifirmaan

 

La BRIE-2 2 0122002

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1M WM“

 

POLITICS OF GENDER DIFFERENCES IN TAIWAN’S LEGISLATIVE YUAN:
DESCRIPTIVE, SYMBOLIC, OR SUBSTANTIVE REPRESENTATION.

By

WAN-YING YANG

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

1999

Professor Mark P. Jones

ABSTRACT

POLITICS OF GENDER DIFFERENCE IN TAIWAN’S LEGISLATIVE
YUAN: DESCRIPTIVE, SYMBOLIC, OR SUBSTANTIVE REPRESENTATION

By

WAN-YING YANG

This study examines the impact of gender on representative politics in Taiwan. I
start my theoretical inquiry by looking at the different assumptions surrounding the
relationship between identity (gender) and representation. Liberal theorists claim that
representatives should transcend any identity difference and move from the localized
concern to the universal interest. whereas the "politics ofdifference" literature posits that
identity characteristics matter and consistently shape deputies’ policy interests. These
two theories thus contradict each other. making opposing assumptions regarding the
existence of a direct and positive link between the large-scale presence of one identity
group (descriptive representation) and the embodiment of the interest ofthat identity
group (substantive representation).

In the face of these competing theories. I test five theoretical hypotheses
regarding the extent to and manner in which female legislators differ from their male
counterparts. The impact of gender in the legislative process is measured on two
dimensions: attitudinal (symbolic) and behavioral (substantive).

In order to test these five hypotheses. I conduct two separate analyses of policy
representation in Taiwan's Legislative Yuan. Two distinct datasets are employed. The
first is based upon a primary content analysis of the bills presented between 1990 and

1997 in the Yuan. The second is based upon interviews with third-term legislators. a

matched sample of men and women. in which I attempted to elicit their views regarding
their policy priorities and the constituencies with which they identify.

Both the quantitative and qualitative findings reveal that female legislators have
developed their own distinctive set of concerns and priorities. Women legislators
embrace such issues as education. health. welfare. children. and women’s issues.
However. male legislators do not share this priority list: rather. they are more likely to
prioritize political and economic issues. The interviews make clear that female and male
legislators also display different constituency concerns. Overall. I demonstrate that

gender has both a substantial and distinctive impact on the political arena in Taiwan.

ACKNOWLEDGMENTS

I am deeply grateful to the members of my dissertation committee. First among
them is Mark Jones. As chair of my committee. he has guided me through the
dissertation process. ensuring I was on the right track and providing me with support to
complete this research. I also would like to thank David Rohde. As a congressional and
institutional scholar. he read through my drafts and made wise suggestions. Two other
members of my committee. Dennis Patterson and Rita Gallin. deserve special
recognition. Both of them made critical comments and compelled me always to
conceptualize the theoretical bases of my work. I am further grateful to Professor Jeffrey
Riedinger. My experience as his research assistant has enriched my research interests.

I also must thank all those in my family for all they have done for me during my
graduate studies abroad. Their encouragement and support have served me
immeasurably in all aspects of my life. Special and greatest gratitude goes to my
parents, Sue-Yuan Yang and Yuh-Chen Hung. for their continuous support during my
study. Last, but certainly not least. I wish to acknowledge my deepest appreciation to
my fiancee. Ju-Hsiou. His abiding love. patience. and engagement during the past years

apart have inspired me to do my best in this project.

TABLE OF CONTENTS

CHAPTERS
1 INTRODUCTION

 

 

 

 

 

 

 

Research Question 1
Theories of Women‘s Representation 4
The relationship between gender and representation 4
The nature of representation 1 1
Literature Review 13
Comparative study on women in representative politics 13
Research of gender dynamics in Taiwan’s representative politics ------- 20
About this Study 25
Organization of the Study 30

 

2 WOMEN IN TAIWAN’S POLITICS

 

 

 

 

 

 

 

 

 

Taiwan’s Sociopolitical History since 1947 33
Political Structure and Institution 36
Electoral clientism 3 6
Electoral rules 38
Gender quotas 4 3
Social Structure and Change 47
Modernization 4 7
Women’s organizations and Movements 5 l
Democratization 54
Democratic transition and women 54

 

The impact of democratization on women’s representation --------------- 57

 

The change ofthe Legislative Yuan 59

 

A Summary 63

THEORETICAL HYPOTHESES AND RESEARCH DESIGN

 

 

 

 

 

 

 

 

 

 

Theoretical Hypotheses 64
The policy domains of gender difference 67
The legislative dimensions of gender difference 69
Behavioral hypotheses 7O
Attitudinal hypotheses 75

Research Methodology 78
Two research methods 78
Data Collection 79
Method 1: longitudinal study of legislators‘ behaviors 81
Method 2: interview study oflegislators‘ attitudes 88

LEGISLTORS [N TAIWAN: A BEHAVIROAL ASSESSMENT

Hypothesis 1: Institutional Placement of Legislators: the committee system -- 91

 

 

 

 

 

 

 

Committee membership by selection 93
Committee chairship by election 97
Hypothesis 2: Participation in Legislative Activities 103
Stage I: Bill initiation 105
Stage 11: Committee chairship l 1 1
Stage III: Speech in the committees l l 7
Stage IV: Speech in the Yuan meeting 122
Summary : Gender difference in legislative participation ---------------- I27

vi

INTERVIEW OF LEGISLATORS: AN ATTITUDINAL ASSESSEMENT

 

 

 

The Interview Data 133

Hypothesis 3: Legislators‘ Perception ofConstituency 135
Factors influencing legislators‘ perception of constituency ------------- 137
Gender patter in constituency perception 141

The recognition of and commitment to represent women‘s concern ---l47

 

Hypothesis 4: Legislators‘ Policy Priority 152

Hypothesis 5: The Association among Different Dimensions of
Gender Differences 163

 

 

A women’s Culture Explanation for the Gender Difference 167

6 CONCLUSION

 

 

 

 

 

Research Findings 178

The Implication and Discussion 184

Suggestions for Future Works 187

APPENDIX 1. Bill classification 190
APPEENDIX 2. Interview cover letter (a) and questions (b) 194
BIBLIOGRAPHY 200

 

vii

LIST OF TABLES

 

 

 

 

 

 

 

 

 

 

Table 1-1. Comparison ofdifferent discourses on the representative democracy ------- 10
Table 2-1. Elected seats (ES) and number of women elected (W.E.) in different

district magnitudes in the Legislative Yuan. 42
Table 2-2. Correlation between district magnitude (DM) and the number of women

candidate elected under the SNTV for the Legislative Yuan election----42
Table 2-3. The electoral outcome in Kaohsiung district under the SNTV

system in 1986. 45
Table 2-4. Proportions of votes and seats of major parties in the Legislative Yuan

election . 1986-1998. 61
Table 3-1. The sample legislators for the behavioral study 83
Table 4-1. Committee memberships from 1990-1997 (3 legislative terms) ------------- 95
Table 4-2. Numbers and proportions of bill sponsorship of each policy area ----------- 106
Table 4-3. OLS models of policy initiation l 10
Table 4-4. Numbers and proportions ofcommittee chairship of each policy area ----- 1 12
Table 4-5. OLS models of committee chairship l 16
Table 4-6. Numbers and proportions of committee speeches made

in each policy area 1 l 8
Table 4-7. OLS models of committee speeches 121
Table 4-8. Numbers and proportions of Yuan speeches made in each policy area ----- 124
Table 4-9. OLS models of Yuan speeches 126
Table 4-10. First policy preferences in different stages of the legislative process ------ 127
Table 4-1 1. Gender difference on the percentage of legislators’ participation in

caring and communication issues 128

Table 4-12 Gender difference on the percentage of legislators’ participation in

viii

feminist issues

 

Table 5-1. The sample distribution by party and gender

 

 

Table 5-2. The sample distribution by age. seniority. and education

Table 5-3. Constituency groups perceived by legislators

 

Table 5-4. Policy priority listed by legislators

 

Table 5-5. Logistic model of legislator‘s policy priority on women’s issues -----------

 

Table 5-6. Women’s groups inside and outside of government

Table appendix 1: Two stages ofbill classification

 

CHAPTER 1

INTRODUCTION

Research Question

To date, the influence of women on representative politics has not been
systematically studied. therefore, one has not yet been able to prove conclusively whether
the presence of women influences political processes and policy agendas to a significant
degree. However, Taiwan. due to its unique history. culture. and political institutions.
offers us a great opportunity to study this question in detail.

In 1947, a formula that reserved for women approximately 10 percent of seats at all
levels of the legislature was incorporated into the Taiwanese constitution. Since the
adoption of this formula four decades ago. the proportion of women representatives at all
levels of the Taiwanese Government has wavered between 10 percent and 20 percent of
the whole. Despite this apparent failure by women to “build” on the seats guaranteed by
the constitution. it can be argued that the system is. nevertheless. acting as an influential
agent of counter-socialization and effectively promoting political representation for
women. This is particularly evident when one considers the cultural and socioeconomic
disadvantages that women face in Taiwan's traditional Confucian society (Chou. Clark,
and Clark 1990).

Pre-1990 studies (Chou. Clark. and Clark 1990) found that women. and particularly
those from elite backgrounds or political families. did manage to overcome cultural and
political disadvantages when entering politics. In addition. it was observed that women
legislators differed little from their male counterparts in terms of role orientations, types
of legislative activities undertaken. and their legislative effectiveness. However, they did
appear to be less active than men in sponsoring legislation. In the light of this. the studies

argued that Taiwanese politics seemed to be closely following the liberal—feminist

tradition. In this comparison, the reserved-seat system is seen to be serving the same
function as the women's movements in America and Western Europe — namely,
overcoming socialization and discrimination barriers by expanding political opportunities
for women.

It is my contention that these accounts are of questionable uses as regards
understanding and evaluating the relationship between women and representative politics
in Taiwan's current political context. Prior to 1987, Taiwan was governed by an
authoritarian regime under which, for the most part, the legislature acted as a “rubber
stamp”. Consequently. exactly who was or was not present during legislative sessions
was largely irrelevant. as the influence of the legislators themselves was limited. Chou.
Clark, and Clark (1990) argued that a "gender-neutral" model best represented women's
political activities before democratization in Taiwan. However, I would argue that these
activities reflected a political reality of homogeneous oppression under authoritarian rule.
Neither gender nor class can be used to differentiate the roles of legislators at that time.
since all were effectively powerless.

Accordingly, studies focusing on the number of women in electoral politics before
Taiwanese democratization can often only result in a somewhat shallow (rather than in-
depth) understanding of women’s representation in Taiwan. since no attempt has been
made to see what gender representation actually means to women. Put another way.
Nechemias's (Nechemias 1994) description of the status of women in the pre-perestroika
Soviet Union. in which the “tokenism” of women’s representation prevailed in a token
institution, even though the proportion of female deputies was relatively high, can in all
likelihood be applied to Taiwan, too. Consequently, without a critical examination of the
nature of the representation. a study focusing on the number of female participants may
well be misleading.

Therefore, bearing all of the above in mind. the aim of this study is to examine the

impact of gender in representative politics after democratization. As political

IQ

democratization is consolidated. and social fissures come to the surface. political
decisions in Taiwan are now subject to the influences of growing demands, conflicting
claims, and limited resources. The questions of “who is present”. “who represents what”.
and “who represents whom” are now of great interest to people studying representative
democracy.

In 1996. Peng, Wan-Ru. a lifelong female activist and a delegate of the Democratic
Progressive Party (the DPP, Taiwan’s first opposition party). put forward a proposal
which provided for a minimum quota of 25 percent of women in the party nomination list
for national elections. The proposal caused heated debates amongst different camps and
was eventually passed after a taxi driver murdered Peng. The death of the female activist.
and the various provisions of the plan itself. provoked nationwide interest in the rights
and security of women in Taiwan. In a direct response to the homicide. a new cross-party
Coalition for Promoting Women’s Rights was formed within the Legislative Yuan
(Taiwan’s national legislature). leading to many of the bills relating to women’s security
being passed immediately. and women’s issues gaining increasing attention.

The institutional change from the national reserved-seat system to the new DPP
party gender quota approach has led to new interest in the notion of women’s
representation. Extrapolating from the experiences of other countries. particularly
Western European and Nordic ones. once a certain level of female representation has
been achieved by one political party, one might expect that other parties would adopt
similar standards (Skjeie 1992). Given the potential impact of gender quotas on women’s
representation, the global debate has long focused on such questions as: why do we need
more women in politics? What differences will result from having more women in
politics? Does gender make any difference in representative politics? And. in what
ways does the policy representation of women occur in the legislative process? The

answers to these questions must be grounded in theoretical and empirical inquiry.

Theories of women’s representation

This study addresses the issue of gender representation in Taiwanese politics via two
inquiries: (1) should gender constitute a significant basis for separate political
representation at all, and (2) how will differences between the genders. if any can be
identified. affect that representation? In addressing the first issue, of whether significant
gender differences in politics exist. and what the nature of these differences might
possibly be. it should be noted that some argue that gender does not make a meaningful
difference at all. However. others argue that women comprise a distinct political group
with distinct political interests that warrant representation. and I will begin by examining
this theoretical disagreement in some detail.

The second issue involves the question of how gender differences. if they exist at all.
are actually carried over into representative politics. Some argue that increasing the
number of female representatives is a necessary step because only women can represent
women. Still others argue that. even with mirror (proportional) representation. the
representation of an electorate’s interests and opinions cannot be guaranteed — i.e.. that
women do not necessarily need to be represented by other women, and that women, when
elected, may not actually achieve anything for, or actively represent. their own sex. This
fundamental disagreement leads me. in the second part. to attempt a clarification of the

nature of representation.

The relationship between gender and representation

Different assumptions are present regarding the relationship between identity (in this
case, gender) and representation. The disagreements are generally centered upon two
questions, namely, what should constitute the basis of representation, and can (or should)

identities (i.e.. gender differences) serve as such a basis? Such questions are the result of

deeper ideological divides between certain conflicting democratic ideals (Squires 1995).l
The concept of liberal democracy has always claimed to transcend any identity
differences, to move from localized concerns to the universal interest. and to deny
identity (gender) as the basis of representation. Nevertheless. discourse as to what should
constitute the content of democracy, and general political shifts in one direction or
another, establish and alter those categories that are seen as relevant and important
enough to be represented. As the notion of identity differences introduces new
perspectives into the democratic discourse. our understanding of the nature of
representation also changes fundamentally. Such “theories of difference” have moved the
question of identity (gender) into mainstream representative discourse, resulting in the
seeking of special forms of representation and political recognition. Relating each to
women’s representation, I will now briefly discuss the two competing forms of
representative democracy - the liberal (individualistic) democracy and the group

democracy.

Liberal democracy: the argument of equality despite difference

The liberal tradition of democracy originates from the ideals of individualism and
freedom from state interference, moving away from group privilege and transcending the
affects of individual difference to work toward an ideal of citizenship (Bryson 1992). At
its centers is a universal norm of citizenship, in which each individual’s level of political
participation and influence is equal. It is held that all people. regardless of their social

and cultural background, should be treated equally and. therefore. that representative

 

' For example, the tradition of liberal democracy is concerned with representing people on the basis of
individual ideas. The notion of pluralist democracy. as advocated by Dahl. seeks to represent different
interests in society via the form of interest groups. The idea of consociational democracy. as articulated by
Lijphart. seeks to represent the social divisions in the form of parties. Finally. the prevailing concern of
democracy of difference is to attempt to represent different identity groups (race, ethnicity. class. and
gender). Here I am especially interested in the notions of representative democracy related to women’s
representation—liberal and difference theories.

democracy does not, or should not. differentiate between citizens on the basis of identity
differences. The central concepts of “the citizen” and “the individual” assume that it will
make no difference whether these citizens or individuals are men or women as
“representation” means representing the abstract individual without distinction. The
liberal tradition has highlighted the virtue of representing people on the basis of
individual ideas instead of social groupings. and interests and ideologies instead of
identities.

The liberal school focuses on individual rights and individual access to political
influence and seeks to “equalize” it by disregarding group concerns and social structure.
To this extent. it is in line with liberal feminism, which advocates equality of opportunity
for women regardless of “difference”. and promotes the entry of women. on the same
basis as men, into established institutions without any concern as to the concept or
premise upon which the institutions in question is based. In seeking increasing numbers
of female representatives. liberal feminists argue that democracy requires this for reasons
of political equality and justice. It is regarded as an injustice that women are excluded
from any form of activity, including politics. and, since women constitute approximately
half of most populations, it is expected that they should be represented more equally than
is currently the case. Furthermore. liberal feminists would introduce the idea of
affirmative action into representative democracy not only to promote equality in terms of
opportunity, but also in terms of outcome: i.e.. in order to achieve the random and even
distribution of political representation.

However, without recognition of the differences between the genders. and of the
need to treat women as a “collective”. a proportionally even distribution of representative
seats by gender will not achieve much more than claims for gender equality in other
occupations have produced. The concept does not address how women represent.
whether they are represented as individuals or groups, or the possible impact of such

representation on women. Neither does it guarantee that these women. once elected. will

make any difference or even represent the interests of women, since both male and female
constituents will have a hand in electing them. It might well address the issue of equality
and fairness re the individual women seeking election. But it does not address the issue
of the representativeness of the candidate or the need of the electorate, as a group, to be
represented fairly (Norris 1996; Phillips 1991; Phillips 1995). Looked at in this manner.
the achievement of equal representation for women would appear to be an end in itself,
regardless whether such women will make a difference. or substantially and effectively
represent the interests of women.

By and large, liberalism assumes that if no obstacle is present to equal participation.
political influence should be distributed randomly and evenly regardless of individual
ethnicity, race, or gender. “The liberal individual is an abstract person without context —
an individual with rights. but without sex, race, age. or disabilities and situated outside
geographical and social boundaries. The ideal liberal society is color-blind and sex-
blind.” (Duerst-Lahti and Kelly 1995, 51) Therefore, representatives should move
beyond any localized or group concerns to embrace a variety of different interests, and
anyone can represent any group’s interests without the need to actually belong to that
particular identity group. In other words, gender identity, or any group identity, is not
seen to constitute a distinct set of interests and values to be represented, rather, identity
characteristics are assumed irrelevant. Despite this. some feminists have pressed for the
politics of difference and heterogeneity to be incorporated into democracy theory and

practice.

The group democracy: the argument of identity difference

Two points of difference between the liberals and the school of identity are. firstly.
that the latter recognizes that women have different interests and form a politically
relevant group deserving of representation, and, secondly, that it considers group

representation an essential prerequisite to the achievement of political equality.

 

.c
I“

 

   

 

 

.
1
.
LI
.
n “

 

 

 

 

 

Some feminists have challenged the gender neutrality of “malestream” political
thought in liberal tradition. accusing it of dishonestly “abstracting” the individual from
the social and economic structures that bring about inequality (Phillips 1993). Whereas
the liberal claim to treat everyone as abstract individuals or citizens. regardless of their
sex. race. or class. is often associated with the ideal of “impartiality” and with the concept
of a “universal” point of view. Young describes impartiality as “an idealistic fiction”
(Young 1990, 104). In her opinion, it is impossible for individuals to adopt a point of
view that is not. to some extent. affected by their situation and, if a point of view is
“situated”, then it clearly cannot be universal. Furthermore. she argues that “because
such a universalist ideal continues to threaten the exclusion of some. the meaning of
‘public’ should be transformed to exhibit the positivity of group differences. passion. and
play” (Young 1990. 97). However. to accommodate group difference. it is necessary to
first recognize the nature of it.

Many feminists assume that women speak “in a different voice”. whether they base
this on biological, psychological. or sociological theories of gender difference (Galligan
1982). Galligan’s influential work, In a Different Voice. has given rise to new theories of
gender that validate “women’s differences”. and have stimulated many scholars from
many fields to explore its implications. Because of sex-socialization. it has been found
that women possess different, but not inferior. moral outlooks to men. Female morality
emphasizes interpersonal relationships, responsibility. and the need to care for others,
whereas male morality emphasizes competitive success, rules, laws. and the protection of
individual rights. Furthermore. adult sex-role differences, as well as the importance of
gender in the structure of women's social and workplace interactions. reinforce and
perpetuate gender differences.

Many researchers, such as Alison Jaggar, Nel Noddings. Susan Okin. Mary O’Brien.
and Sra Ruddick (Rhode 1990). have also stressed the importance of care-giving

relationships in explaining attributes historically linked with women. Most feminists

agree that the production/reproduction divide ascribes different social roles and functions
to men and women, and that women’s continued responsibilities as the primary care-
givers of the family, lead to women assuming different positions. characteristics. and
interests to men. As regards politics, all these differences are expected to lead to a
distinctive female perspective on many issues. and to the nurturing of different political
interests and behaviors to men (Kelly. M.A.. and Horn 1991). According to these studies.
women will initiate. pursue, and support issues concerning bio-social production and
reproduction, i.e.. those issues that concern the control of. responsibility for. and care of.
both people and other natural resources.

Building upon this recognition of gender difference, some feminists propose new
visions of democracy based around the concept of heterogeneity. seeking to recognize
and accommodate identity difference by promoting group representation (Phillips 1991).
They insist that one’s gender identity also determines one’s claim to be representative of
one’s constituents. The politics of difference. as advocated by Iris Young (Young 1990).
posits that existing arrangements are unjust and unrepresentative, since they fail to reflect
the diversity of the population. Young argues that groups that have suffered oppression
need “group representation” to ensure their inclusion and to diminish the effects of
“impartiality.” The politics of difference is an approach that claims to involve all groups
by prescribing different and specialized treatment for oppressed and disadvantaged
groups, and its treatment of the issue of female political representation involves such
recognition of group difference. As Young asserts. “social justice requires not the
melting away of difference. but institutions that promote reproduction of and respect for
group difference without Oppression.” (Young 1990, 47.) Therefore. it is argued that only
through such a form of group representation as this can the democracy achieve an
effective level of representation of the distinct voices and interests of those oppressed
constituent groups.

In a similar vein. Charles Taylor in his The Politics ofRecognition (Taylor 1994)

criticizes the “difference-blind” liberal theories as a reflection ofone hegemonic culture.
He argues that the concern for equality should involve the recognition of the value of the
diverse cultures and identities and, in order to attain the required recognition. an
increased public presence of the “oppressed” is necessary. He argues that, given the
diverse group perspectives and concerns that exist. the polity can only function
democratically by recognizing and incorporating the voices of these different groups. For
example, by bringing more women into politics. we bring a different (and differing) set of
values, experiences. and expertise to our politics.

In short, the differences between these two claims are derived from their different
premises. i.e.. the concept of equality despite difference versus the concept of equality but
difference; and opposing viewpoints as to just how such a difference will alter and affect
representative bodies and public policies. Table 1-1 is an attempt to summarize the
different meanings of “women’s representation” within each democracy discourse. as

well as related implications and goals.

Table 1-1: Comparison of different discourses on the representative democracy

 

 

 

 

 

 

 

 

Democracy Premise Unit of Representation Goal

discourse representation Implication

Liberal Individual Individuals Equal representation as an Discount the

democracy equality end itself (politics of idea) difference

Politics of Group Group Group representation as a Accommodate

difference difference means to justice (politics of the difference
presence)

 

Liberals attempt to equalize political power with equal representation without
emphasizing the difference in the representative process and outcome. whereas the latter
view, that of women as a distinct group with shared basic interests, implies that the
“individual” of liberal theory does not. in fact. exist. According to the former, it is not
necessary to have any group representation since identity is not necessarily linked with
interest, whereas. according to the latter approach. group representation is necessary to

bring gender difference into politics by linking the identity with interest. Staying with

 

this line of reasoning, I will investigate the following questions — Are there significant
gender differences in politics? Will an increase in female representation change policy
directions and existing political institutions? And. finally. how can such a transform

proceed and on what basis?

The nature of representation

Having claimed the right to representation as a distinctive group. the next question
to face women concerns the form such representation should take. A variety of possible
approaches to female representation have already been mentioned. Other women who
share similar characteristics with them may represent women. or people whom they
believe to represent them, or even people who share similar concerns and opinions with
them but whom are not necessarily women. In light of these different approaches to
representation, Hanna Pitkin (Pitkin 1967) distinguished three types of representation:
namely, descriptive. symbolic. and active. Descriptive representatives refer to those who
share the same characteristics as those represented; symbolic representatives refer to
those who symbolize the identity or quality of a particular constituency; and active
representatives refer to those who are concerned and act for the interests of those
represented. According to these different types and definitions of representation, women
politicians who stand for women either descriptively or symbolically may not act for
them. and those who act for them may not necessarily be women. Pitkin further criticizes
descriptive representation by arguing that. because the notion of who is present in
legislative assemblies diverts attention from what the legislators represent, fair
representation is achieved only by looking at legislative behavior - by focusing on the
activities of a person rather than their characteristics. Pitkin argues that even the
maximization of descriptive representation, such as via a huge numerical increase in
female representatives. may be of limited help to female interests because these

representatives may not effectively represent women’s interests and ideas.

Perkins and F owlkers (Perkins and F owlker 1980) make similar distinctions of
representation but under different rubrics. with “social representation” being similar to
descriptive representation. and “opinion representation” being similar to active
representation. In a simulated choice situation, Perkins and Fowlkers found that both
men and women would prefer to have their interests or ideas represented rather than
achieve mere social representation. They also agreed with Pitkin’s theory that
representatives could appear to be a microcosm of society in every aspect but still fail to
represent what the electorate wants. Conversely. legislators could do what the electorate
wants without resembling their characteristics. Put another way. it remains unclear
whether the election of more women candidates will bring about a more “women-
orientated” political agenda.

Ann Phillips (Phillips 1995) also offers a similar framework of representation by
distinguishing “politics of idea” from “politics of presence”. The politics of idea
emphasizes “what” is represented by looking at the activities and decisions regardless
who is present in the legislature. whereas the politics of presence emphasizes “who” is
represented by looking at the composition of a parliament. However. Phillips places
greater emphasis than Pitkin on the importance of the latter approach, questioning
Pitkin’s assertion that the emphasis on what they represent must depend primarily on a
tighter mechanism of accountability that binds politicians to the opinions they are meant
to represent. Phillips. as concerned with accountability and the search for authenticity as
other feminists. goes so far as to question who can best stand for another.
Acknowledging that representatives have considerable autonomy. as well as the absence
of procedures for establishing what the groups want and think. Phillips still believes that
who those representatives are is more important than what they represent.

However, Phillips assumes that we can better approximate true representation by
ensuring that representatives are a microcosmic reflection of the group they are meant to

represent. Therefore. in the case of female representation. because they stand for women

(symbolic representation). it is expected and assumed that they will act for women
(substantive representation). But this argument is based on the assumption of a direct
correlation between gender (presence) and interests (idea). In reality, women politicians
are divided by other social differences — class, race. age, and marital status — that might
very well prevent them from simply “standing for women”. In addition, they are also
forced to confront, and perhaps grapple with. pre-existing institutional norms and
priorities and will, like their male counterparts. have most likely come into politics
through party systems that may separate them from each other as regards basic ideologies
and priorities (Skjeie 1992). In short. there are many pre-existing fissures between
women that may significantly affect the use of the female mandate.

Pitkin’s theoretical discussions about different types of representation, and Perkins’
and F owlkers’ evidence, all confirm that only women can represent women descriptively.
But all of them question whether women officeholders would necessarily represent
women substantively, acting for women and their interests. This assumption leads them
to emphasize the importance of the “politics of idea”. In the other corner. Phillips
assumes that there is a link between descriptive and substantive modes of representation.
and therefore argues for the “politics of presence”, believing that the presence of more
women in political office is necessary for the effective representation of women to be
achieved. Clearly, then. the fundamental disagreement between the two “camps” rests on
one unanswered question: namely. whether women in office act for, or simply stand for.
women. Or to put it another way: Are female representatives more likely than their male

counterparts to actively represent women?

Literature Review
Comparative studies on women in legislative politics
The theoretical dispute regarding whether. and in what way, representation is related

to gender identity raise a number of questions. Do women make a difference in politics?

13

More specifically. under what conditions can women in politics make a difference? And.
in what ways do women make a difference? Though much research has been undertaken
in an attempt to answer these questions. the results have been mixed. 1 have classified
these past comparative studies into four distinctive groups. (1) Those researchers who
have stressed similarities between male and female politicians. (2) Those who have
identified distinctive female policy interests and agendas. (3) Those who have suggested
that gender differences in politics depend on the institutional context in which they exist
(critical mass theory). And. (4). those who have suggested that gender differences are
contingent to lots of different factors (contingent and contextualized theory). These four

distinct “sets” of research findings are summarized in the following subsections.

1. Similarities between male and female politicians

Early studies (Gehlen 1977; Kirkpatrick 1974; Mezey 1978; Muller 1982) reported
that women tended not to identify themselves as “women’s candidates” and. therefore.
did not place a higher priority on women’s issues than did men. A possible explanation
for this is that women politicians may wish to avoid being labeled “radical”, or only
responsive to the interests of women, fearing that voters and male colleagues would
penalize them for this. Using attitudinal surveys and statistical analyses of voting
behavior, Carroll, in Women as Candidates in American Politics (Carroll 1985), noted
that most women candidates did not belong to feminist organizations and did not
emphasize women’s issues. Diamond (Diamond 1977) came to a similar conclusion.
noting that the majority of female state representatives in the United States did not
campaign on women’s issues or trumpet feminist goals. In a study of women in a US
State legislature, she found that only a small number of the female representatives were
advocates of women’s rights. Rather. most could be classified as the traditional civic
worker, whose attitude toward the role of women in society was one that assumed that

little or no change was needed. These studies starkly demonstrated that female legislators

l4

..‘

 

v.
.,‘
»
‘ n
u
u
u
, I
h
x
'9.
‘.
.‘. -
n w
n
h

did not distinguish themselves from male legislators as regards “women’s concerns”.

In addition to their role orientation, Randall (Randall 1987) argued that. due to the
process of selective recruitment. party pressure, and political socialization within
legislatures. female MPs maintained similar political attitudes to their male counterparts.
Clearly. these earlier studies. carried out in the United States. suggested that the claims
that women politicians would transform politics were overstated.

Studies in other countries also generated somewhat similar results. In the United
Kingdom Parliament. with few women entering politics in the early 19705. women MPs
were loath to adopt exclusively women’s issues. feeling that their loyalty was to their
party rather than to women (Vallance 1979). Comparative studies of female leaders, such
as Benazir Bhutto. Corazon Aquino and Margaret Thatcher. found it difficult to classify
these female leaders as a single group as regards both their leadership styles and policy
agendas (D'Amico 1995). Recent research has also uncovered an increasing similarity
between Scandinavian female and male MPs in their methods of political communication

(Karvonen and Selle 1995).

2. Differences between male and female legislative participation

Contrary to these studies. however. some researchers had found gender differences
in legislative attitudes and behaviors. In terms of legislative activities. some researchers
have shown that female legislators are far more supportive of women’s rights and are
more likely to sponsor bills that address matters of health and welfare. and women.
children, and family (Carroll. Dodson. and Mandel 1991; Reingold 1992; Thomas 1991;
Thomas 1994; Thomas and Welch 1991). Scandinavian research also uncovered a
gender-related division as regards parliamentary activities. In Dahlerup (Dahlerup 1988)
and Haavio-Mannila’s (Havvio-Manilla 1985) study of Nordic politics. women MPs were
more likely to ask questions about family. health. housing. environmental. and consumer

policies, while men were more likely to ask questions re fiscal. labor market. and energy

15

 

 

v- )‘H'

’ i\

 

 

   

   

policies. In addition. women were most likely to raise the profile of women’s issues on
the legislative agenda.

There is also some evidence to indicate a gender gap in the basic political attitudes
of representatives. In Reingold’s (Reingold 1992) analysis of the survey responses of
state legislators in California and Arizona. she found that female politicians were more
likely to represent women’s interests. to perceive themselves as receiving support from
women constituents, and to consider such support important, than were men. In addition.
women politicians felt themselves to be more qualified to deal with the concerns of their
women constituents. Moreover. Lyn Kathlene's study of female legislators in Colorado’s
House of Representatives (Kathlene 1989) suggested that female representatives
conceptualize problems differently from their male colleagues. Thus. she argued that
women tend to practice a different type of politics to men. and have different conceptions
of what constitutes political power.

In terms of political ideology. Stanwich and Kleeman (Stanwick and Kleeman 1983)
noted that women officeholders tend to be more sympathetic to liberal policies such as
nuclear disarmament and welfare programs. It has also been found that women are more
liberal than their male counterparts as regards their general attitudes and legislative
behaviors. although the differences are not great (F rankovic 1977: Gehlen 1977; Leader
1977; Norris 1986; Welch 1985). Similar attitudinal gender differences have been found
amongst the US party delegates and state legislators (Baxer and Lansing 1983; Diamond
1977; Soule and McGrath 1977). Moreover. such results also gain support from other
studies carried out in other countries. For example, studies in the United Kingdom
suggest that women are more left wing and supportive of feminist values than men are
(Norris and Lovenduski 1995).

With respect to legislative duties and roles. some studies also point to gender
differences in terms of a legislator's own self-perception. A legislator's perception of

their individual role shapes what they do in the office and how they do it. Norris’s study

of British MPs (Norris 1996) distinguishes three types of legislators: the constituency
workers, the parliamentarians. and the party loyalists. It was found that women,
primarily because of their greater emphasis on interpersonal relationships. their self-
perception as problem solvers and public servants. and their feelings of obligation to the
communities they represent. tended to give a higher priority to constituency work
(Antolini 1984; Freeman and Richardson 1995; Thomas 1992). Kirkpatrick (Kirkpatrick
1974) argued that women legislators tend to see themselves as problem-solvers, in power
in order to help people. whereas male legislators tend to see themselves as leaders or
politicians. It should also be noted that female legislators receive more calls from
constituents and spend more time serving constituents than their male colleagues do.
Thomas’s study (1993) of city council members produced a similar result, and Freeman
and Richardson's (1995) survey of four state legislatures confirmed that. after controlling
for other variables, women received more casework requests than their male colleagues.

and were twice as likely to believe that they performed more casework.

3. The critical mass theory

Given the fact that these various sets of findings seem to contradict each other, other
studies have sought to explore the conditions under which the similarities and differences
between female and male legislators might become more apparent. Some have argued
that once women achieve a “critical mass” of political positions of power, they would
start to change politics in substantive ways. The concept of critical mass is borrowed
from physics. where it is used to refer to the quantity needed to begin a chain reaction
(Dahlerup 1988). The theory as it applies to representative politics suggests that political
behavior is shaped by its surroundings (loosely. a new “institutionalist” perspective). and
that politicians will respond differently and strategically to the opportunities that form
around them. According to this theory. few gender differences in legislatures will be

apparent so long as women remain a distinct minority. because of the possible trade-offs

l7

between challenging the dominating organization's agendas and being marginalized
politically that could take place. However, the situation may be transformed once
women reach a critical mass (Dahlerup 1988; Kanter 1977; Saint-Germain 1989; Thomas
1994). However. the proportion of the total number of seats which constitutes a critical
mass vary, from 15% to 30% (Dahlerup 1988; Kanter 1977)

For example. Saint-Germain’s (1989) longitudinal study on Arizona state
legislatures from 1969-1986 found. firstly. that women legislators were more likely to
introduce feminist legislation seeking for greater equality and improvement of women
status, as well as legislation dealing with women traditional interests such as children.
education. and the family. In addition. as the number of women representatives
increased, the degree of attention paid to women issues also increased.

In addition to the longitudinal study. Thomas and Welch (Thomas and Welch 1991).
in their twelve-state survey of state legislators. also found a correlation between the ratio
of female delegates and the degree of support for legislation related to women, further
confirming that the proportion of women present makes a difference. Although Thomas
and Welch did not find a systematic sex difference, they did find that women tended to
introduce more bills relating to children. women. and the family, and that they were more
likely than men to cite such bills as accomplishments. This was particularly so in states

with a high percentage of female deputies.

4. The transformational and contingent theory

Nevertheless, recent research indicates that the “number factor” alone is far from
adequate in explaining whether or not women make a difference in politics. Yoder
(Yoder 1991) argues that simply increasing the number of women in the workplace does
not solve the negative effects of tokenism, since token men do not experience the same
level of inequality and negative treatment as token women. The problem of tokenism is

sexism, and it is the group norms rather than the group size that brings inequality.

18

Hellevik and Skard (Bystydzienski 1995) applied Diamond’s classification in their
study of female Norwegian local councilors in seven municipalities in the county of
Akershus. They found that. even though women occupied approximately 30 percent of
the seats at the local level in Norway. only one-third of the women representatives were
women’s rights advocates who “identified” with women and their claims. and took
initiatives on women’s issues and interests. It is clear that, without a shared
consciousness of a women’s agenda. female politicians would not take any action on
behalf of women even ifa large number of women representatives ever come to sit in our
legislative bodies. F urthermore. the attempt to shape politics and legislation to reflect
women’s interests does not necessarily guarantee that the outcome of such political
processes will have the intended consequences. Clearly. whether women politicians will
make differences depends on various factors.

It is possible for women to make a difference in politics when their numbers are
relatively small, as in the case of Japan (Bystydzienski and Lin 1994: Ling and Matsuno
1992), and to seem to have little effect when their numbers are increasing, as in Canada
(Brodie and Gotell 1991). In the former Soviet Union. women made up a fourth to a third
of the political representatives. yet made no significant difference at all. Recent studies
(Epstein 1983; Reskin 1988; Yoder 1991) indicate a curvilinear relationship between the
size of a minority grouping and institutional responsiveness. That is. very small numbers
of the minority receive a hostile response from the dominants. but. as the size of the
minority increases. the response from the majority becomes more positive. However. as
the number continues to grow. the majority turn full circle. reacting with hostility to what
they now perceive as threatening. This finding suggests that the presence of a larger
number of women in politics does not make it easier for women politicians to develop
and promote women’s agenda. It is also worth noting that, the larger the number of
women, the more divergent women’s opinions may become. especially when female

politicians are scattered across different parties and have embraced conflicting ideologies.

Bystydzienski (Bystydzienski 1995) concludes that whether or not women have an
effect on politics cannot be determined solely by the number of women in the legislature.
Rather, the effects of their presence depend on other factors. such as the political climate.
the prevailing ideology. the strength of the women’s movements. and the structure of the
political system. When the political climate is favorable to the entry of new groups. the
social and women’s movement flourishes. and the political structure is relatively open

and flexible. women’s interests and concerns may finally change the face of politics.

Research of gender dynamics in Taiwan’s representative politics

As noted above. many researchers in many different countries have studied the
representation of women. However. the majority focus upon the industrial West — such
as the United States. the United Kingdom. and certain Scandinavian countries. Though
some studies were undertaken in developing countries, they were not as reliable as the
former in terms of quantity and quality. The reason for this concentration on the politics
of the West is twofold. Firstly. these countries are well-institutionalized nations with a
long history of parliamentary politics. Even with different constitutional systems (the
parliamentary versus the presidential system). the legislatures within these countries do
wield important influence in the policy-making process. Not only that. but the function
and operation of the parliaments in these countries has been stable over a long term and.
because of this relative institutional stability. systematic data is available for research. It
is especially true of the United States that congressional studies have developed plenty of
solid tools (e.g., interest group ratings. and indices) to analyze roll call data and
legislative processes. Secondly. a social interest in gender politics has also been
apparent. Most of these countries have either the highest levels of female representation
(like the Scandinavian countries). or a history of well-organized women’s movements
(such as in the United States). leading to a social climate that is supportive of such

studies. For example. in the United States. a critique of academic bias toward gender

20

studies (Sapiro 1981) provoked a response from certain academic institutions. In the
APSA and MPSA, a women’s section hosting various gender study panels was created.

On the other hand. in developing countries. if there is any interest at all in studying
the role of female politicians. it is likely to be focused on the executive side of
government, given the usual administrative domination over the legislature. The situation
in Taiwan is similar to this. In general. Taiwan’s legislative studies have reflected the
political structure and the processes ofchange. Before democratization (1987), most of
the research merely provided static descriptions, and few behavioral studies were
available, since the legislative body was not an autonomous stage in the legislation-
making process. Put simply, the behavior of individual legislators in an ineffective
institution was not worth studying.

Chou, Clark, and Clark’s work (1990) represents the first systematic analysis on
female political representation in Taiwan. They were primarily interested in assessing the
problem of women’s political under-representation in Taiwan, and did so by conducting a
survey of the representatives in 1985. They pointed out that several institutional and
social factors combined to produce certain opportunities and barriers to women’s political
participation in Taiwan. Within the legislative bodies. they found that women entered via
reserved-seats systems and differed little from their male counterparts in terms of role
orientations, types of legislative activities, and legislative effectiveness. with the
exception that they were less active than men in sponsoring legislation. However. despite
the well-stated theory. methodology. and findings. the study was conducted in 1985.
when Taiwan was still under the influence of an authoritarian regime. At that time, the
representative bodies and members played an extremely limited role in the decision-
making process. Taiwan has experienced democratization since 1987 and, therefore,
most of the study's findings must be re-evaluated in the light of this.

The number and scope of Taiwanese legislative studies remained limited even after

democratization. What research does exist in this area focuses either on the legislators’

21

.-

interpellation actions or on several particular “women's bills.” Basically, the legislative
study of gender politics is something that has arisen only in the last ten years. and the
quantity and quality of relevant research is still limited. Cheng’s (Cheng 1983) analysis
of legislator's interpellations on women’s issues. from the 67‘h to 71St legislative sessions.
found that gender and seniority were the most significant variables. In other words,
women and senior legislators were more likely to raise questions in the policy area of
women’s interest than male and junior members were. Nevertheless, as a whole, those
relating to women’s issues made up only a tiny minority of all interpellation.

Own-Ying (Own-Ying 1992) conducted a content analysis on representatives’
interpellations and legislation on women’s issues from the 515’ to the 84'h legislative
sessions (17 years). She found that the frequency of interpellation on women’s issues
was very low. and that more were put forward by female legislators than male ones. In
addition, there were more such interpellations by supplementary legislators than senior
legislators, and by the Taiwanese than mainlanders. Consequently, it can be seen that
gender, seniority, and ethnicity were the main factors that dictated the frequency of
interpellation on women’s issues. Similar content analyses of legislators’ interpellations
(both oral and written) were conducted by the New Times F oundation. from the 79‘h to
the 82nd legislative sessions. The findings were similar to previous ones. namely that
gender and seniority are still the most important factors in explaining the difference in
interpellations on women’s issues. but that the result is not overly significant.

Most of the above studies on gender difference in Taiwan’s legislative politics were
concentrated upon such interpellation. which is but one function among various
representative activities. The benefit of such an approach is that there is a considerable
volume of relevant data. Prior to democratization. because of the possession of only
limited legislative power, most legislators regarded their duties as monitoring what
government was doing instead of making laws. Therefore, interpellation had been the

primary channel by which representatives could express their opinions about how the

Ix.)
IQ

Government conducted its business. Besides. interpellation is a near individual action.
and consequently unlike other legislative actions. such as bill sponsorship and debates.
which involve cooperation and bargaining between legislators. It is easy for an individual
legislator to proclaim his or her concerns regarding any specific topic by questioning the
administrative officers on their policy-making or policy implementation.

However, the system of interpellation was not as institutionalized in Taiwan as it
was in other parliamentary countries. In Taiwan’s Legislative Yuan. if legislators are not
satisfied with the replies they receive from administrative officers. they cannot find other
ways to gain satisfaction. because they do not have access to other leverages. like a power
of veto. investigation. or impeachment with which to do so. In addition. and as most
legislators admitted. their assistants usually write the actual contents of an interpellation.
Legislators under the SNTV electoral system (discussed later) are constantly striving to
secure the attention of the media and thereby raise their personal profile. To achieve this.
they will inevitably pick the most “popular” subjects upon which to interpellate.
depending on the news. social events. or current topics of interest to the media itself.
Hence the fact that essentially redundant interpellation is extremely common. Many
scholars, as well as some legislators themselves. have criticized the pattern and function
of interpellation. However, due to re-election concerns, legislators must compete to show
off. even if the issues upon which they choose to focus are of little or no interest to them
at all.

In addition, the establishment in 1995 of the Forum ofNational Affairs, an hour-
long meeting that precedes the Yuan (floor) meeting. has created another stage for
legislators to express their opinions on general policies and social developments. It has.
in fact, replaced some of the functions of interpellation itself. The channels through
which legislators can express their opinions have multiplied and overlap. but the
substantive function and policy influence of these simple opinion expressions remains

limited. The studies focusing on interpellation for understanding legislators’ interests and

23

behaviors are thus problematic.

Learning from the experiences and methods of American congressional studies.
different approaches to legislative studies have emerged in the 19903. Huang’s study
(Huang 1994) of legislators’ constituency services found that women, on average.
received 200 casework requests from their constituents. while men received only 112.
The result of conducting a multiple regression on the second term legislators. and after
controlling for their party, education. seniority. backgrounds. and constituency
geography. proved that gender was a significant factor in explaining the variation in
representatives’ casework. This result is consistent with some past U.S studies referred to
earlier (Antolini 1984; Freeman and Richardson 1995; Thomas 1992). However, Huang
also noticed that at that time there were only 17 women legislators. constituting a mere
10.6 percent of the whole. and that. of those 17. only 15 agreed to be interviewed. Also.
most of her female interviewees were either highly recognizable to the public, or had
previous experience as elected officials with good local connections. This could be
another reason for the high number of casework requests they received from their
constituents.2

Cheng (Cheng 1997) also conducted a thesis on legislative participation in women’s
issues. Learning from Sheng’s sophisticated study (1996) on legislative behavior in
Taiwan, she concentrated upon bill sponsorships and speeches to examine the possible
gender effect on representative politics. In the three women’s bills Cheng selected. she
recorded the frequency of sponsorship, co-sponsorship, and speeches of each legislator

serving in the second and third terms. and applied OLS regression to the participation of

 

2 In the beginning, 1 also attempted to ask the relationship between legislators’ role definition and
constituency service in the interview. However. most legislators I have contacted did not have ideas about
the approximate number of case services they received and the content of the requests. and most of them
have their assistants take care of the constituency service. Most legislators’ assistants know more about the
constituency service than the legislators did. However. few assistants did sort out the types and the number
of the legislators’ constituency service in a week or month, on average. A general survey in the number of
the case services should be careful about the mis-reporting or over-reporting.

24

the legislators in each bill. For those three bills. she concluded that three independent
factors - gender, connection with women’s groups, and serving as committee chairs — are
significant in determining the intensity of a representative's legislative participation in
women’s issues. Her study attempted to combine the strength of legislative study with
gender study; however. her choice to focus on three particular feminist bills cannot serve
to illustrate the role of gender representation in legislative structures and processes as a
whole.

About this study

The results of these comparative empirical studies on gender representation have
been contradictory. Also. the researches carried out in Taiwan. which either applied
loose measures like interpellation or restricted itself to certain select and limited cases.
are not as solid and systematic as the studies carried out in other countries. Drawing on
the experiences and insights outlined in the existing research. and using both comparative
legislative studies and Taiwanese studies, I attempt to retest the empirical basis of the
gender difference argument in Taiwan. and address some of the questions encountered in
previous research designs. My research shall address some of the problems and enrich
the study of gender representation in the following aspects:

F irstly, gender difference in politics does not exist in a vacuum. Gender is one
dimension among various sociopolitical divisions (class, party. and ethnicity). and it is
dangerous to assume that an essential clear-cut difference will always be apparent
between men and women. Policy environments. both political and social, determine the
possible policy-making processes and potential outcomes of any political initiative. Even
differences in the institutional environment can lead to different representative outcomes.
A number of previous studies were conducted in the legislature across different cultures
and political institutions. and the results were markedly different. An understanding of
Taiwan’s unique sociopolitical context is an important first step in any attempt to explain

its gender dynamics. Therefore. in my study. I first turn to the contextual factors.

Ix)
UI

political as well as social. structural as well as concerning processes. and examine the
ways in which they might affect the gender dynamics in politics.

Secondly. most of the studies emphasize only a single dimension. focusing on either
the attitudinal or behavioral side of political representation. Unfortunately. when
focusing upon the behavioral side. one looks at policy output or roll call data to determine
the pattern of gender difference. which does not take into account legislators’ subjective
willingness to represent. or the institutional constraints imposed upon women who are
willing to “push” women's issues. Therefore. this approach tends to underestimate the
gender differences in a more subtle way - perceptual and symbolic. On the other hand.
focusing on the perceptual dimension can. potentially. overestimate gender differences
and lead to findings that ignore or obscure reality. Only a few studies (e.g.. Thomas
1994) have examined legislators’ perceptions and behaviors. Clearly, a study
incorporating the perceptual input from legislators and the policy output might better
explain the gender dynamic in representative politics.

To address the bias of studying a single dimension. my research is designed to
investigate three dimensions. (1) The objective legislative behaviors of representatives:
By examining everything from agenda setting to policy deliberation from l990-1997. I
will show the general trend and pattern of gender representation in Taiwan over time. (2)
The subjective perceptions of female representatives: By interviewing third-term
legislators, I will reveal the “inside story” of gender dynamics in the Legislature. (3) The
links between subjective identification and objective representation: By linking both
aspects together. I will be able to explain and put in context the gender politics of
Taiwanese representation in a thorough manner. In short, I will specify the dimensions
(both behavioral and attitudinal) that need to be addressed if one is to successfully
analyze the policy representation of women. and I will apply different methods of
investigation to each of these “dimensions”.

Thirdly. the choice of appropriate measurements for legislative behaviors is crucial

26

to the study of policy representation. A mature gender study on representative politics
requires not only that the researchers possess a sophisticated understanding of the theories
of gender politics. but also that they are aware of. and can comprehend. the workings of
the Legislature. Gender study assists researchers in formulating their questions within a
theoretical framework. and legislative studies are essential in so far as they allow the
researcher to approach these questions with a sufficiently in-depth understanding of the
legislative process. Therefore. in order to offer a fair account of gender representation in
Taiwan. my study incorporates the strengths of both gender study and legislative study.

In previous behavioral research in this area. some methods of measurement have
proved problematic from the perspective of both gender and legislative study. As
mentioned above, one such problem is the use of parliamentary interpellation statistics. as
this measure cannot directly address the policy representation of legislators. Of course,
the most common approach in use is that of examining roll call data, however. many
researchers have also cautioned against the use of voting statistics as an indicator of
gender differences.

Tamerius argues that roll call voting is “the indicator theoretically associated with
the least dramatic aspect of gendered representation.” (Tamerius 1993, 96) She claims
that the reason why previous studies of gender differences in legislative politics have
failed to find compelling evidence of a significant female impact is because they relied on

roll call voting statistics. She contends that.

Roll call voting. which assumed that enactment is the most important stage in the
legislative process. privileges majority and. therefore. male interests. Since policies
of concern to the majority of members are bound to make it to the floor eventually.
early maneuvers are unlikely to have a major impact on whether a policy is
ultimately adopted. From the perspective of women and other legislative minorities.
however, critical stages of the legislative process are more properly identified as
agenda setting and policy formulation, since the vast majority of policies of interest
to underrepresented groups. including feminist bills. never receive consideration on
the floor (1993: 105).

She further argues, both theoretically and empirically. that the greater the level of

27

 

“purposive” involvement required by an activity. the greater the difference between male
and female legislators in terms of the performance of that activity in feminist issues. She
ranks the extent of these sex differences in the following manner: roll call voting
(smallest difference): co-sponsorship: speechmaking: and sponsorship (largest difference
between the sexes).

From the perspective of legislative study. it is also noted that the representatives
engage in various legislative activities other than voting on bills in order to pursue their
concerns. For example. legislators might set the agenda and specify alternatives. exclude
some alternatives from consideration. initiate or draft a bill or an amendment, debate for
or against the policy. or attempt to persuade legislative colleagues to adopt the same

position. As Hall has noted, if we neglect to take into account the various possible efforts
that legislators might put into the legislative process before reaching the stage of formal
roll call voting, we might well underestimate their overall representation. Voting records

alone are not sufficient. In Participation in Congress (Hall 1996), Hall cautioned us that,

Although most members vote when specific decisions come to a formal roll call on
the chamber floor, floor voting is only one and probably not the most important
form of participation in the legislative process. Building a coalition for a legislative
package. drafting particular amendments. planning and executing parliamentary
strategy. bargaining with or persuading colleagues to adopt one’s point of view—all
these activities weigh more heavily than voting in the decision-making calculus of
most bills, especially in a legislature where committees, anterooms, and staff
meetings are the principal forums for deliberation and decision. In such activities.
participation is highly selective. . .(1996: 8)

In light of the selective nature of legislative participation and the different
assessments of level of representation. therefore. a detailed and systematic investigation
of legislative activities is clearly necessary. Among the various legislative activities that

I will identify, my research will emphasize the importance of certain stages of the law-
making process in Taiwan. from agenda setting to policy debates. to an understanding of
political gender differences. I will also compare the varying degrees of gender difference

that can appear at different stages of the legislative process.

28

 

P'.) .qu

\
L‘Auub.

'\
..,-‘ A: e .

l".1*‘ V

‘ .
1.....uiu.

.
Iv v

 

 

 

 

 

. J:
\M‘
‘ I ~
'~_ 75.
,u‘
‘.'4
no ‘ .7
a.

F inally. in an attempt to bolster the strength and reliability of their legislative
research, most recent studies of T aiwan’s legislature borrowed their methods of
measurement from the experienced and systematic research previously carried out in the
United States. However. researchers should be aware of the inapplicability of some
measures employed in traditional legislative studies to the case of Taiwan. In an
American legislative study. Hall (1996) argued that the structure of levels of legislative
participation in Congress is hierarchical. and that. from the top. the order was: engaging
in agenda action. offering substantive amendments. offering minor amendments of
procedural motions, speaking during markup. voting and merely attending. and. finally.
engaging in none of the activities. Members carrying out activities at the top of this
hierarchy were likely to engage in lower level activities. too. Another warning in this

area comes from Kingdon (Kingdon 1989). who notes that voting is a legislative activity
that actually requires little substantive knowledge. since representatives are able to rely
on cues from a variety of sources when making their decisions.

However, Sheng’s study of Taiwanese legislative politics (Sheng 1996) questioned
Hall’s notion of this hierarchical order of legislative activities. The problems of using
Such “American methods” to measure activity in Taiwan are twofold: firstly, there are
difficulties re the availability of data (e.g. voting statistics); secondly. one must allow for
the different nature of the legislative process in Taiwan. In Taiwan, because of
inSfitutional limitations (e.g.. quorum requirement on initiation), and the ways in which
the parties mobilize legislators and direct their activities. legislative participation is no
10Tiger hierarchical but. rather, can be described as random.

To devise a set of measures applicable to the study of Taiwanese legislature requires

a Contextual understanding of the operation of Taiwan’s Legislative Yuan. Drawing on
previous and related studies. and incorporating the results of my interviews. I will devise
a set-of measures that will be of use in gaining a full understanding of the workings of the

Taiwanese legislature. I also intend that this will assist in the study of gender

qu .
h

a 'J\

 

A.“ i

 

 

”A

 

1‘1”

 

 

representation in this arena. This aspect will be addressed and further elaborated upon in
the research method section.

In closing. I hope that such a study of gender representation in Taiwan will enable us
to project, accurately but within the bounds of reason. the possible consequences of more

female officeholders for both public policy and political processes.

Organization of this study
With the aim of examining the gender dynamics present in Taiwan’s legislative
politics. this study is divided into two parts. In the first part. I examine the impact of
contextual factors - such as sociopolitical structure and change - on the political status of
women in Taiwan. Following on from this. I turn to the impact of gender difference on
representation in Taiwan's national legislature. I examine this by. firstly. analyzing the
possible behavioral differences between male and female legislators in the Legislative
Yuan between 1990 and 1997. and. secondly. exploring the subjective perceptual
differences between the genders in the third term-legislature.
In pursuit of the above. I attempt in Chapter 2 to explore the former of these two
Parts. Three dimensions that are related to the issue of political opportunities for women
to enter politics are especially addressed in this chapter: political structure and institution.
SOcial structure and change, and democratization. When referring to political structure
and institution. 1 mean the constitutional arrangements, electoral cultures, the SNTV
e1ectoral rule. and the reserved-seats provisions in Taiwanese politics. I also provide a
bI'ief account of the evolution of the political institutions within the existing social
Culture, in so far as this is relevant to the issue of female political opportunity and
Participation. Moreover. levels of female political participation are sensitive to the
Clynamic characteristics of societies. making it imperative to consider the change in a
Social context. With respect to social structure and change. I pay particular attention to

Ihe role of social modernization and the women’s movements in changing the political

status of women in Taiwan. F inally, I address the relationship between democratization
and female representation and, in doing so, I attempt to specify both the positive and the
negative impact of contextual factors on female representation. All of these structural
descriptions will set the stage for the following chapters.

In Chapter 3. I place my main research questions into a set of theoretical hypotheses
and research designs. The main theoretical purpose of this study is to explore the gender
dynamic and to thereby specify the nature of women’s representation in Taiwanese
politics. As mentioned earlier. embedded in the theoretical debates are two questions —
(1) Do women make a difference in Taiwanese politics? (2) If so. in what ways does
gender make a difference? As regards the first question, an expectation that women
participate differently in legislative process arises from the assumption that women
deputies would better represent female concerns. Therefore, differences as regards policy

representation between male and female deputies are also assumed to lie in this female
area. Based on the existing research and theoretical projections. I define the gender-

Specific policy domains - “women’s issues” - to which female legislators are expected to
Pay more attention than their male counterparts.

The second question touches on the nature of gender representation. asking whether
the women in legislative politics are making a difference in a substantive way or merely
On a symbolic level. It is my intention to explore how different modes of representation -
Clescriptive. symbolic. and substantive - are linked together in Taiwanese politics and just
how representation can be seen to embrace both the attitudinal (symbolic) and the

behavioral (substantive) modes. An awareness of both “dimensions” is crucial to a
Comprehensive understanding of gender representation. Therefore, I generate five
hBipotheses concerning gender difference along the lines of these two modes of
representative politics. and follow this with a research design employing these two
distinct methods of assessment. I then test the five hypotheses - by means of a behavioral

aSsessment in Chapter 4. and an attitudinal assessment in Chapter 5. Therefore. Chapter

4 focuses on the actual behavioral differences between the genders in legislative politics.

and Chapter 5 examines the perceptual differences regarding legislative roles and

priorities between female and male legislators.
Chapter 4. in focusing on the actual behavioral differences between the genders.
seeks to investigate the latitude ofsubstantive policy representation of deputies in the
Legislative Yuan. To explore the trends in legislative participation. I conduct a
longitudinal analysis of legislators’ behavior in Taiwan’s Legislative Yuan from 1990-
1997. In studying legislative behavior in this manner. different aspects of legislative
processes are also examined. from static institutional arrangements to dynamic legislative
processes. As regards the former. I examine the division of labor in terms of legislators’
committee memberships in the Legislative Yuan. Turning to the dynamic aspect. I
observe the legislative process from agenda setting to policy formation. I start by looking
at the agents of agenda setting. examining the initiating and scheduling of bills. Next, I
examine the policy formation process. by examining the speeches that the legislators have
made in the Legislative Yuan. Having divided legislative processes into these distinct
and recognizable stages. I then go on to examine possible gender differences at each
Sta ge.
As already outlined above. in Chapter 5 I focus on the subjective aspects of
I'epresentation. By examining the perceptual differences between the genders. we can
better understand the mechanisms behind gender patterns — patterns that reveal
thernselves in legislative activities and behaviors. In this chapter. I am especially
itlterested in individual legislator's perceptions of what. and whom. they regard it as
inlportant to represent. In other words. the extent of their identification with their
cOnstituency and their policy priorities. The interview is regarded as the most direct
method with which to reveal legislators’ conceptions of representation and, therefore. I

design several questions to elicit their views on these two aspects of representation.

Finally. the interviews are of a matched sample of third-terrn legislators.

b)
[J

CHAPTER 2
WOMEN IN TAIWANESE POLITICS

Taiwan’s sociopolitical history since 1947

For centuries. Taiwan's social structures have preserved both Confucian and
patriarchal traditions that have relegated women to a subordinate social status. These
traditions have excluded women from active roles in the public domain. including
politics. However. the area of female political representation in Taiwan reveals an
interesting contrast. On the one hand, women have fared well in Taiwanese politics
compared to women in other countries with similar cultural backgrounds (Chou, Clark.
and Clark 1990). On the other hand. however, many factors have hindered an increase in
the number of female representatives over the past decades. Thus. viewing the situation
from a dynamic perspective. the proportionate representation of women remains static
even though the sociopolitical environment has changed. To better explain and
understand the opportunities as well as the constraints on female political participation.
this chapter will explore women's political status in Taiwan by examining it within the

Context of change in the sociopolitical structure.
Between 1947 and 1997. Taiwan has experienced a significant change in its

Sociopolitical structure. In 1947. the Kuomintang (the KMT. or the Nationalist party) lost

the civil war in Mainland China and moved to Taiwan. The regime the KMT established.

Which survived intact for almost four decades (1947-1987) was essentially an
al-lthoritarian one-party government composed of mainlander political elites. The party
I'llled Taiwan by implementing martial law. which repressed competing political parties
and any political activity.

Upon retreating to Taiwan in 1948. the KMT formally maintained a five-branch

r1'dtional government. including the Executive Yuan (the administration). the Legislative

Yuan (the parliament). the Judicial Yuan. the Examination Yuan. and the Control Yuan.
Nevertheless. informal political practices and institutions greatly affected the operation of
the formal institutions of government. especially so given the fact that the KMT has
dominated normal political practice at all levels of government for forty years. The KMT
held almost all the significant administrative positions and exercised a major role in
policy making. with the major policy decisions being made by the top party organs
(Central standing committee). The executive Yuan was viewed as an organ that
implemented rather than initiated policy, and members of the Legislative Yuan (national
parliament) were known simply for endorsing the decisions made at the party center.

The KMT also retained a four-tier administrative and representative system designed
for all of Mainland China. from the national down to provincial. county/city. and
town/village levels. At the national level. in order to maintain the claim to be
legitimately representing all the people of China. the KMT kept those representatives of

the Legislative Yuan elected from China in place for a long time without making them
subject to re—election. Yet. the increasing age of the incumbents as well as the pressure to
make the electoral organs more representative of Taiwan’s constituency led the regime to
hold “supplementary” elections in 1969 in order to "rejuvenate" national representative
DOIitics. The supplementary elections elected only a small minority among those tenured
members from China before democratization took place in the late 19805 (Chu 1994). At
the local level. because of the earlier ethnic conflict (the 228 event) between the
mainlanders and the Taiwanese in 1947. local elections were instituted in the 19505 so as
to reduce the level of Taiwanese hostility toward the mainland KMT regime. and to avoid

the threat of local violence. The native Taiwanese were able to elect and enter

representatives up to the provincial level. and executive officials up to the county/city

1eVels. These elections allowed the native Taiwanese to enter local politics and thus

Prepared Taiwan's citizens for future democracy.

In addition to maintaining political stability. the KMT also embarked on a process of

34

 

 

 

4'”

IM.

4 r

u
l
I

‘h.

 

modernization in seeking industrialization, urbanization. and economic prosperity. For a
long period. the KMT sustained political stability with economic growth under the
authoritarian regime. By the 19805. Taiwan’s political landscape had. however. changed
dramatically. With the increasing political protests and social movements, Taiwan began
the process of democratization with the lifting of martial law in 1987. The first
opposition party, the DPP (Democratic Progressive Party). was legally established in
1988. and the 1989 legislative election became the first nationwide campaign after
democratization. Since 1989. the DPP has consistently won approximately 30 percent of
vote in the national representative campaigns. and the electoral share of the KMT has
been steadily eroded during the transition from a one—party authoritarian regime to a
competitive party political system. In the early 19905. the entire Legislative Yuan and
the top executives became subject to election. and party competition was further
intensified in 1993 with the split of the New KMT Alliance from the KMT to organize
the third party. This party was termed the New Party (NP). and their electorate is heavily
concentrated in urban areas among mainlanders and young middle-class voters (Tien
1 996). However. despite the KMT’s decline. they still remain the majority party in
Taiwan.
For the past five decades (1947-1997). the paralleled engineering ofelectoral
POI itics, social development. and democratization has carried important implications for
developing an understanding of women’s representational politics in Taiwan. In the pre-
1 987 authoritarian phase. women who entered politics were the result of the dynamic
interaction between political institution and social modernization. To ensure
SOeiopolitical stability. the regime informally utilized electoral clientism in order to
tighten social control. detrimentally affecting, in the long term. the opportunities open to
Women to develop an independent political status (Chen 1996). The regime also formally
implemented the SNTV electoral rule and certain gender quotas. which have helped as

“’63“ as hindered the potential growth of Taiwanese women politicians. The social

35

program of modernization. which ensured Taiwan’s economic prosperity. also altered the
existing gender social structure. though in a more positive direction. By 1987. the
democratic transition, which has restructured the political scheme of Taiwan. has brought
another wave of women into the domain of representative politics. However, over time.
women in Taiwanese politics have experienced changing political opportunities and
fluctuating constraints derived from the interaction of institutional design and structural
reshuffling. Therefore. in the following sections. I examine the implications of the

electoral culture and system, of the social structure and change. and ofdemocratization.

to women’s representational politics in Taiwan.

Political structure and institution

Electoral clientism
Since the evacuation from Mainland China to Taiwan. the KMT has implemented
local elections at the level ofcounty magistrates. city mayors. city and county
councilmen. and provincial assemblymen. As an extrinsic regime to most native
Tai wanese, the KMT controlled the elections by co-opting the existing social structure
and local elites by cultivating networks of patronage. By exchanging and allocating
economic interests and local rents to local factions. the KMT was able to mobilize local
electoral support with which to bolster its political dominance. To strengthen its political
Control, the KMT pursued bi-factionalism. i.e.. they kept two local factions competing for
Pelitical offices through divide-and-rule tactics. Candidates were mostly from local
factions and were prohibited from engaging in activities outside their own voting
districts, so much so that an opposition coalition could not be formed. The KMT
therefore gained its legitimacy by fostering institutionalized connections between newly
arrived mainlander elites and local elites through a patron-client network (Tien 1996).
- The KMT’s patron-client mechanism is operated through ties with local factions.

Which have the strongest webs ofconnection. Especially in traditional rural areas.

36

everyone is connected to every one else in a complicated web of social relations. These
faction webs. or networks. are made up of many types of connections - kin connections
(family, relatives. and patrilineal and filial ties). neighbors. classmates. co-workers. or
other bonds. The kinship and patron-client nature of these electoral politics greatly
influenced the operation of local politics and women’s political power at the grass roots
level in Taiwan.

Within these connections. women are relatively disadvantaged. since the women’s
network is less extensive and their status is more subordinate to the patrilineal or filial
relationship. Women have ties with their family. their other relatives. and with groups of
women with which they socialize. most often in their neighborhood or their natal villages.
However. women leave their natal family upon marriage. and hence leave their original
network of connection. transferring to become a subordinate of her husband’s network.

M en, on the contrary. following Chinese convention. live in the same village in which
they were raised. Therefore. by comparison. men can maintain continuing and broader
Vi l lage networks (Bosco 1994. 124) in which their kinship relationships can be
COntinuously built up or maintained.

Therefore, the political power women can accrue is limited due to women’s

d i scontinued connections and subordinated relationship in the traditional network. The
e l eetoral culture in general is a disincentive for independent women to enter politics. and
Ihen almost always occupy the leadership positions within the factions. Only under rare
COtiditions (e.g.. the reserved-seat system, discussed later). do they recruit relatives or
Otlrler loyal women to run for political offices. Women politicians that have arisen from
the political factions or families are generally seen as fulfilling the need for a political

' ' heir". Their qualifications as political leaders are questioned. since their success

d epends on the political capital of their families. This type of woman politician does not
1'1 ecessarily succeed in politics. and their political careers are tied to their family elders’

( 11’) ale normally) agreements and party’s support which might undermine their subsequent

37

ability to gain true independence and political stature.

Chiang’s interview (Chiang 1996) found that even though female politicians from
factions or political families might not start their political careers voluntarily. they could
still manage them successfully given their advantages in temis of political socialization.
sufficient training. and political resources. Also. because of their factional backgrounds
and political experience. their views of power and politics are more realistic and
skeptical. However. the patronage orientation of the local networks and factions do limit

a woman’s ability to act independently and to deal with the social problems of special

interest to women (Clark and Clark 1997).

Electoral rule
Different electoral systems also affect the electoral prospects of women and various

minority groups. Taiwan employs the single non-transferable voting system (SNTV).
like Japan’s Lower House before 1994. The majority of Taiwan’s legislators. both local
and national. are elected under this system. in which votes are cast for only one candidate
in a multi-member district. and candidates who get relatively more votes than others win
the seats.

Generally speaking. women did better in such multi-member districts with

pro portional representation than in single-member. plural systems. Willma Rule (Rule
1 9 87) found that party-list proportional representation (PR) systems averaged a 12.5
percent female membership in their lower houses. while other systems averaged only a
4 - 1 percent female membership. Furthermore. her multivariate analysis showed that the
e: x i stence of proportional representation with a party list was the major determinant of the

l evel of women’s representation in these 23 democracies. controlling for other

3 Ocioeconomic and political characteristics.
~ Comparing four parliamentary—plural democracies with nine parliamentary—PR

C Q untries in terms oftheir performance records regarding female representation. Lijphart

ni‘

. z

(Lijphart 1994) also found significant differences between these two systems. He noted

that women legislators were more numerous in PR countries than in plural countries.
Engstrom (Engstrom 1987) examined the relationship between district magnitude and
female representation in the Irish Dail. His findings suggest that districts with four or
five seats are more likely to have female representatives than districts with only three
seats.

The prominent explanation for differences in terms of women’s representation levels
among the different electoral systems is the district magnitude. Some recent research
(Darcy. Welch. and Clark 1994; Matland 1992; Rule and Zimmerman 1994; Studlar and
Welch 1989) also found that more women were elected under multi-member districts than
under single-member districts. The reasons why women’s representation is advanced in
multimember district with proportional representation are various. From the parties’

perspective, under a PR system. parties have more incentive to place different based

Candidates, like women. in the same list to broaden their appeal and to run as a team. At

the same time. the potential cost to the party of nominating a woman also decreases. The

po 1 itical risk of nominating women decreases as the district magnitude increases. because

unl ike the situation that occurs in a single-member district system. a female candidate did

1101: exclude all male candidates in a multimember district. From the voters’ perspective.

V'Oters may be more comfortable voting for women as part ofa slate of representatives

than as their sole representative. From the female candidates’ perspective. they might

fTeel themselves to be more acceptable to parties and voters, and thus be more willing to
run for election. In general. a multi-member district system can lower the barriers for

Women, 50 they would be willing to stand for election. be acceptable to the party

3 e l ectorate, and be approved by the voters. Thus. the election turns from a zero-sum

g arne under the plural system into a positive-sum game under the multi-member district

3 y stem.
The institutional arrangements and legislative consequences for SNTV countries.

such as Taiwan and Japan. are a little more complicated than for a simple PR or plurality

system. Unlike the plurality system. under which the winner takes all and minorities run

a higher risk of losing in a district. the SNTV is not a zero-sum game in terms of party
nominating. Unlike the PR system. under which the party puts the name list together and
allocates the votes to its candidates. under the SNTV, the individuals who get the most
votes win. Even with multi-member districts. since voters can only cast one vote in the
ballot. many voters would still be unwilling to cast their sole vote for a woman as their
representative (Amy 1993; Berg 1994; Chou. Clark. and Clark 1990; Darcy. Welch. and
Clark 1994; Matland 1992) (Studlar and Welch 1989). In terms of minority
representation, the SNTV system does manage to avoid the disadvantages of winner-
takes-all plurality systems. but it does not have multiple votes and a party list as in multi-
member districts. PR systems. Therefore. though the SNTV could be seen as less

attractive than straightforward PR. it does appear to be preferable to the plurality system.

at least for its multi-member district aspects. Consequently. the SNTV system is

I“ e garded as an intermediate stage between plurality systems and PR systems. both

because of its electoral allocation rules (which are close to the plural design). and its

d i Strict magnitudes (which resemble the PR system).
Rule (Rule 1987) compared the consequences ofelections held at the same time but

under different electoral rules, in Japan. In Japan, elections for the Upper House were
he 1 (I under the PR system. and those for the Lower House were held under the SNTV
Sy stem (prior to 1994). The electoral outcome in the SNTV multimember districts was
Very similar to that of single-member districts. since voters selected only one candidate
and the district magnitude was, on average. small. Rule argued that the SNTV system for
tlle Lower House discouraged the nomination of women by major and minor parties since
tlie stakes are so high. Alternatively. the PR system employed in the Upper House turned
0 ut to be more far more favorable to women’s representation. A similar conclusion is

a 1 so reached in Taagepera’s finding (Taagepera 1994) that. although women and

40

 

LL

1..

 

minorities benefit from the increased district magnitude of PR-electoral systems. the
SNTV is as unfavorable to women and minorities as a single-member district, plural
electoral system. Hickman’s most recent findings (Hickman 1997) also showed that
increasing district magnitude under Japan’s SNTV had a negative effect on the election of
women.
However, this conclusion is rarely tested, except in the case of the Japanese Diet
Lower House. In Taiwan. the SNTV system has not discouraged women’s political
participation as in the case of Japan. Women legislators average 10-15 percent of the
members elected since the 19805 and Taiwan has achieved the most equitable degree of
female legislative representation of any country with an SNTV electoral system. The
percentage of female representatives in Taiwan is significantly above the worldwide
average for national legislatures and is close to that of Western European countries with

multi-member PR electoral systems (Chou. Clark. and Clark 1990).

In Taiwan. the districts are based on the existing executive unit and the magnitude
Varies across different districts, depending on the population of the executive unit.
Before the 1989 election. with limited elections at the national level. there were only a
few districts (9), and most districts (7 out of the 9) were large ones with at least five seats.
Thus it is difficult to determine if increasing the size of the district affects the percentage
0 15‘ women getting elected. However. since the district was redrawn at the 1989 election.
i t has produced a variety of district magnitudes ranging from one to seventeen.

Therefore, examining the effect of district magnitudes on the election of women in the

L. egislative Yuan finally becomes possible.
Table 2-1 displays the change in the district magnitude and number of women

e lected across time in the Legislative Yuan election. As it shows. the number of districts
C hanged from 9 to 27 in 1989. and the district magnitude thus began to vary greatly after
the. 1990 election. Table 2-2 examines the correlation between the district magnitude and

Women’s representation under the SNTV system. and this correlation is both positive and

41

Table 2-1. Elected seats (ES) and number of women elected (W.E.) in different district

magnitudes in the legislative Yuan.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Magnitude 1983 1986 1989 1992 1995
E.s./ W.E. E. s./ W.E. E. s./ w. E E. s./ w. E E. s./ w. E
1 1(11/0 1(1)/ 0 8(8)/ 1 5(5)/ 0 5(5)/ 0
2 2(1)/ 0 2(1)/ 11 1ot5y0 10(5)/1 10(5)/ 1
3 9(3)/ 2 9(3)/() 9(3)/ 1
4 24(6)/3 12(3)/1 12(3)/ 2
5 10(2)/ 2 10(2)/1 5(1)/ 1 10(2) 5(1)/ 0
6 6(1)/ 1 6(1)/ 1 12(2)/3 18(3)/3 24(4)/ 4
7_ 21(3)/3 14(2)/ 1
L 16(2)/ 1 8(1)/ 2 8(1)/ 1
1 18(2)/ 4 18(2)/1 I8(2)/2 18(2)/ 6
_1_0 10(1)/2
1:15 11(1v2
goo 16(1)/2 17(1)/ 2
Igtal 53(9)/ 8 55(9)/7 79(26)/ 12 119(27)/ 12 122(27)/ 18

 

 

Note: figures in parentheses are the frequencies of each district magnitude.

'I‘able 2-2 Correlation between district magnitude (DM) and the number of women
Candidate elected under the SNTV for the Legislative Yuan election.

Year Numberof Number

0/o of female Correlation Correlation between

women of MP5 MP5 elected between DM and DM and number of
MPs elected elected by by SNTV number of district district female
by SNTV(all) female candidates delegations
__ g _ SNTV(all)
1 992 12(17) 119(161) 10% .891** .746**
295 18 (23) 122(164) 16% .672** .658**

 

N Ote: the figures in parentheses indicate the total MP5 including non-territorial members.

*3]:

correlation is significant at 0.01 level.

significant. Districts with several seats available are more likely to attract a large number
of women candidates to run for election. F urtherrnore. the district magnitude is also
positively associated with the chance of women getting elected in the SNTV electoral
system. This assertion holds true in Taiwan’s SNT V system and strengthens the
argument for adopting larger district magnitudes in non-PR electoral systems.

The recent ongoing debate in Taiwan about whether to adopt a single-member
plurality system to replace the existing SNTV system is therefore expected to alter the
existing electoral course in many aspects. including that of female representation. In the
interviews. some female legislators did express concern that the change from the SNTV
to the single-member plurality system might be harmful to them. One senior female
legislator strongly opposes the single member plurality system because she claims that it
might totally deprive women of their political opportunities (interview: 98031). Another
KMT female legislator feels uneasy about redrawing the magnitude ofdistricts so as to
create smaller ones, since she considers it a threat to her chances of re-election to be
forced to fight a smaller district (interview: 98002). It appears that some legislators.

especially female legislators. are aware of the effect of electoral rules and district

magnitude on the future prospects of women both seeking re-election. and those looking

to enter the Yuan for the first time.

Gender quotas
Another major reason for the relatively high level of women’s representation in

Taiwan compared to Japan lies in the system of reserved seats. Chou, Clark. and Clark.
attributed the differences as regards female representation between the Japanese and
Taiwanese legislatures under the SNTV system to the use of the reserved-seats system

(gender quota) in Taiwan. By the 1947 constitution, women are guaranteed some seats in
Taiwan’s legislative assembly. In most of the elections. women run in the general

constituencies. and female candidates are guaranteed a certain minimum number of the

seats in each district. The exact number of guaranteed seats and the mechanisms for
electing women are left to the statutory law and are, therefore, slightly different at
different levels of the representative bodies. In the Legislative Yuan election. one seat is
reserved for women in a district with a total of five to ten seats and another reserved seat
is added whenever the district magnitude is increased by ten. In practice, the number of
seats reserved for women has never exceeded one per district, since Taiwan’s district
magnitude has never exceeded 19. An electoral district needs to elect 20 members before
a second reserved seat is designated for female candidates.

There are two mechanisms that allow women to be selected from these reserved
seats under the SNTV rule. Firstly. when a sufficient number of female candidates win.
they are viewed as having fulfilled the quota reserved for women. However, when none
or not enough female candidates are running in multimember districts. the votes of all the
women candidates are tallied separately, and then the female candidate or candidates with
the highest number of votes are elected to fulfill the seats reserved for women (Chou,

C lark, and Clark 1990). For example. in 1986 legislative election, in Kaohsiung district,

there were five seats and I 1 candidates were running in that district. Under the normal

SNTV rule, the first five candidates with the most votes won the seats. However. when

the reserved seat is taken into account. it is mandatory that one out of five seats must be

assigned to a female candidate. The electoral result was changed due to the reserved seat.
Table 2-3 shows the electoral result. In this example. the district magnitude equals five.
therefore. one reserved seat was assigned for female candidate No. l l. and male
Candidate No. 8 was excluded even he got more vote than No. 1 l.

The formula for calculating the quota possesses a markedly “U-curve” distribution
instead of being proportional to the population share of women. When the district
magnitude (total number of seats) is small. for example 5. the proportion of women’s

guaranteed representation appears to be high around 20%; as the district magnitude

increases from 5 to 19. the average proportion of women’s reserved representation

44

Table 2-3. The electoral outcome in Kaohsiung district under the SNTV system in 1986.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Candidate Sex Votes Vote ranking Elected
nee. Ming-chu M 2.892 1 1 No
fihang. Chun-hsiung M 77.429 3 Yes
TChou, Kao-chuan M 4.805 10 No
4. Dai. Chung-ching M 13.378 7 No
5. Su, Chiu-cheng M 11.184 8 No
6.Wang. Yi-hsiung M 91.984 1 Yes
7, Ma. Wei-pang M 5.724 9 No

TChang. His-yao M 54.896 5 No (excluded)
9. Huang. Cheng-hsiung M 62.159 4 Yes
10. Hsiao. Chu-chiao M 86.903 2 Yes

 

11. Wu. Der-Mei F 35,747 6 Yes(reserved)
Source: The Electoral Bulletin of Central Election Commission.

decreases to 5% before it increases again. Only when the district magnitude reaches 20
will the level of representation reserved to women increase again to a maximum of 10%
(Chou, Clark. and Clark 1990). Given the fact that the district magnitude in Taiwan’s
legislative elections range from 2 to 17. the reserved-seats do not perform the same floor
function at all levels of district magnitude. The low level of representation for women
prescribed by law only aims to set the minimum threshold for women’s political
representation.

Generally speaking. the purpose of the quota system is to promote political equality
between the sexes. It is widely recognized that the gender quotas create incentives for
parties to nominate women as legislative candidates. For a period of time. the reserved
seat system had clearly been beneficial in so far as it was promoting women’s political
representation in Taiwan. The proportion of women in Taiwan’s assemblies is quite high
in relative terms. especially in light ofthe island’s patriarchal society and level of
development. Indeed. the reserved-seat system motivates parties to recruit women
Candidates to run for the reserved seat. for fear of an opposing party filling it. Once in
public office, female representatives. who were first elected as tokens by the reserved-

seat, develop their own political skills and are eventually transfomied into competitive

45

candidates. Over time women clearly did become competitive in their own right as
evidenced by their winning significantly more than their minimum number of reserved
seats. During the 19705 and 19805. women consistently won more local council positions
than they had been guaranteed. At the provincial and national levels. women’s ability to
win more than their minimum quota grew much slower (except for the Taipei City
assembly and National Assembly). Yet by the mid-to-late 19805. women were able to
exceed their minimum quotas in the Provincial Assembly and Legislative Yuan as well.
However. attributing the success of female levels of election to the reserved-seat
system is no longer such an unquestionably valid approach. Since the late 19805. the
female share of the seats in most assemblies has grown only marginally. Despite the
advantages women representatives may have attained through this institutionalized
guarantee over the past four decades. as the sociopolitical context has changed, this
institutional arrangement has become more ofa ceiling which stagnates the growth of
women representatives rather than a floor which assures a certain level of female
representation. Since 1987. Taiwan has entered a new stage of socioeconomic
development and democratic transition and consolidation. More eligible women
candidates have moved forward. yet, parties still appear reluctant to nominate more than
one woman in each district beyond the level prescribed by the quota (Chou 1995).

It is also the case that. at the moment. almost all of the female candidates win in
their districts without the need for their gender quotas. Clearly. the quota is too
conservative to produce any progressive effects on the level of women’s legislative

representation. Nowadays. in Taiwanese politics. almost none of the establishment
Women are elected through the reserved-seat system, as evidenced by the claims made
during my interviews with the third-term legislators. the majority of whom won first or
Second place in their districts. Women have gradually become independent actors in the
sehse that they win enough legislative races beyond their reserved minimums to

demonstrate that they are viable candidates in their own right. Some of them even

46

mentioned the negative effects of reserved-seat system as regards its influence on a
party’s nominating and electoral campaign strategy. One senior female legislator
reported that her campaign consultants told her to target another woman candidate instead
of other male candidates as her primary male opponent. She disagreed with that strategy.
but the situation certainly does arise. However. as a whole. she supports the system of
gender quotas due to her recognition that not many women possess the sort of social and
familial resources that she does (Interview: 98031).

In some interviews. women legislators suggested a more progressive measure of
gender quota to kick start female representation out of its current state of slumber. Most
of the female legislators do not espouse the Nordic European model of relatively high
quotas (30-40 percent). instead. they suggest an increase in the size of the gender quota
from the current level of approximately 10 percent to something approximating to a
quarter of the seats available. They realize that elite women can hardly understand the
needs of women of other classes. and the needs for a higher gender quota, which would

both encourage women to run for election in the first place. and force the parties to be to

nominate more women.

Social structure and change
Modernization
The changing gender roles in Taiwan should be assessed against a backdrop of

Confucian patriarchal culture and an ongoing modernization process. As well as
attempting to maintain the traditional culture and level of social control. the KMT also
Sought social modernization. committing itself to promoting rapid economic development
and industrialization and to taking care of all groups and classes through legal reforms
and universal education (Chu 1994). Consequently. during the most recent decades,
TaiVvan has experienced a dramatic economic expansion known as “The Taiwan Miracle”

(RUbinstein 1994). However. those social changes that have accompanied this process of

47

modernization. have brought tension to a social structure deeply rooted in a patriarchal
culture.

On the one hand, the traditional patriarchal culture prescribes women’s roles as
being complementary to those of men, and women are expected to serve different
functions that assist with the overall development of society. In the 19605. educated
women were encouraged by the government to participate in the work force to fuel
economic expansion (Gallin 1984). Women were seen as an efficient and flexible reserve
army of labor with which to bring about economic prosperity but. as economic prosperity
contributed to the growth of a new urban middle class in the 19705. the government then
began to promote the wife and mother role as the proper sphere for women to adopt

(Farris 1994).

On the other hand. this process of modernization contributed to the change in
women’s roles and still helps. indirectly. to enhance women’s political status. In theory.
socioeconomic progress as a result of modernization should break the traditional
patriarchy and kinship culture and enhance the status of women in society. In addition.
the rising educational levels provide women with the means to realize their potential.
move into the workplace. and perhaps alter their social status. Industrialization creates
opportunities for women to leave the home and enter the workplace. A5 a result. the
increase in employment can endow women with material independence, increase their

self-esteem. and give them a degree of control over their lives. Moreover, women’s
entrance into the labor market moves them from the private to the public spheres. changes
the traditional socialization pattern. and fosters their interests in public affairs.

And in practice. the mandated primary and secondary education and the merit-based
access to higher education in Taiwan did create an enlarged group of well-educated, well-
adj usted. women. With respect to the average years of schooling, in 1951, men had more
than Women (3.9 to 1.4 years). but by 1983. this gap had shrunk considerably (8.9 to 7.4

)- AS regards college education. women’s share of this grew from 18.8 percent in 1958 to

48

43.6 percent in 1987 (Chou 1994). This improvement in women’s educational standards
also laid the foundations for their equal entrance into various occupations. The rapid
industrialization in Taiwan shattered the traditional labor model of“men go out. women
stay in” by bringing more women into new jobs in various industries. service-based
sectors. and professional occupations. From the 19605 to the 19805. women have
increasingly participated in the expanding industrial production sector and have
significantly succeeded in gaining entrance to the more desirable clerical and professional
jobs on almost equal terms with men. In twenty years. modernization had altered the
social gender structure in term of education growth and occupation distribution.
Consequently, the increase in education and employment has led to a change in
women’s political participation. In the past. in terms of voting behavior. women were
considered as more conservative or dependent than men. and more likely to be influenced
by their husbands or parents in voting decisions because of their family roles (Hu and
Chen 1987). Women were more likely to vote because of relational connections. like
friends, classmates. and neighborhoods. Also women voters were more likely to be
mobilized by factions: female factional voters (59.6 percent) outnumbered male faction
voters (40.4 percent) (Chen 1994). In terms of party preferences, it was found that the
majority of women voters did not have any (52. 5 percent of women compared with 45.8
percent of men). Among those female voters with party preferences. only 8.8 percent of
Women voted for the DPP; whereas 16 percent of men voted for the KMT (Huang 1994).
In fact, in general. it was found that the majority of women voters supported the KMT,
and that the DPP’s electoral support was drawn largely from male voters. This particular
gender gap would seem to indicate that women tend to be more conservative and thus
more supportive of the KMT.
The gender gap as regards political participation is in many ways decreasing
(improving) as a result of women’s increasing education and employment opportunities.

In addition, the conservatism of women voters was found to be related to their lower

49

education, profession. and family status. Some scholars stratified voters by education.
profession. and marital status. and found that the differences between men and women
decreased within the same stratum (Fan and Hsu 1994; You 1989). As women voters
have obtained higher levels of education and forced their way into better professions. the
gender gap has clearly narrowed. proving that women’s political conservatism and
dependence was caused by their education and marital status rather than their gender. A
comparison of female voting patterns between the years 1983 and 1992 found that the
turnout of women’s votes had increased over time. and women had gradually been
politically mobilized - i.e.. they'd gained higher political interests. formed clearer party
preferences, and had cultivated higher democratic values.

However, despite these obvious improvements. one must still describe the impact of
modernization upon women’s political status as mixed. Whereas all the above is
certainly true. and women's passive forms of political participation (such as voting) now
exhibit a far higher level of sophistication: in terms of active political participation (e.g..
participation in campaign activities). the gender gap does not yet appear to have narrowed
much (Chao 1995). Under the KMT's one-party rule. women’s political role was
substantively limited and the percentage of female party cadres in the KMT serves as an
indicator of just how insubstantial women’s political power under the KMT was. Though

past research found that the educational level of KMT’s party cadres increased over time.
it was also the case that the percentage of women cadres decreased during the same
period. For example. whereas at the provincial level, the KMT cadres with college
education increased from 53.4 percent to 76.7 percent. at the county level. from 38
percent to 64.9 percent, and at the district level. from 40.2 percent to 85.3 percent, the
Percentage of female cadres within the KMT did not grow over the same period. In fact.
female cadres decreased from 8.0 percent to only 5.4 percent at the district level. and
from 17.4 percent to 15.9 percent at the county level. Only at the province level did the

percentage increase slightly (Dickson 1996). To put it simply. in electoral politics as a

50

I-J

whole growth in the level of female representation did not keep pace with changes in the

socioeconomic structure.

Women’s organizations and movements

Another aspect of social change that is important to women’s political power lies in
the evolution and development of women’s movements and organizations. Viewed by
most Taiwanese as a foreign regime. the KMT reoriented the social organizations to
accommodate existing social structures and tight social control. In addition to the
electoral clientism exhibited at the grass roots level in order to co-opt the existing
sociopolitical network into the party structure. in the social sphere. the KMT applied
social corporatism to restructure civil society and voluntary organizations.

On a national level. the KMT divided society into functionally specialized groups.
labor unions, farmer associations. fisherman associations. trade associations. intellectuals.
students. and other professional groups. Before the democratic liberalization of the late
19805. there was only one official recognized union or association within each section.
Any other paralleled organizations were prohibited. The KMT controlled and
demobilized all modern social sectors through the pre-emptive incorporation of all social

groups. and. because these associations were closely controlled by the KMT. their
members had little or no autonomy. It has been argued that this social corporatism had
successfully controlled and atomized society, and that. as a result. Taiwanese civil society
had been deprived of the option of autonomous collective action (Chen 1996; Chu 1994).
Since tight control was exerted over the non-governmental organizations under
martial law. the vast majority of women’s groups were either controlled by the KMT.
SUC h as the Chinese women’s anti-communist aggression league. and the F u-Nu Huei
WOmen’s associations. or were local branches of conservative international organizations.
such as the YWCA. the Zonta C lub. and the Taipei Jaycettes (Farris 1994). Within these

organizations. women served as wives ofcelebrities or maintained their own social status

within the traditional and conservative value system that dominated. Most of the
women’s organizations maintained a close association with the KMT and followed the
policies decided by the Office of Women’s Activities of the KMT central party. under the
leadership of Chiang. Sung. May-Ling. Autonomous women’s organizations promoting
social reform did not exist until the 19705.

The earliest attempt to form an influential women’s movements in Taiwan dates
back to the early 19705. Lu, Hsiu-Lien, the pioneer of Taiwan’s women’s movement.
launched the organization with a talk on International Women’s Day. 1972. Under the
authoritarian regime. she undertook a series of lectures. debates, and writings to raise the
consciousness of women. However. her first attempt to organize a women’s organization
- the Association for the Promotion of Women - was thwarted by the Taipei City
government in 1972. One year later. in 1973. the Taipei Chapter of the International
Federation of Business and Professional Women. aiming at sponsoring activities and
networks for career women, was formed. However, the founder, Lu. was later forced out
by the wives of KMT officials. At the same year. the Protect You Hotlines. which were
set up to counsel and help women who suffered domestic violence or sexual assault. were
closed down within two years ofopening - sabotaged by the KMT. The feminist

publications distributed by the Pioneer Press (established by Lu) were censored and
denied a copyright by the government. The first stage of the women’s movement was
regarded as a one-person crusade by Lu (Lu 1994). and. given the extent to which it was
attacked. this was no doubt the case as women rushed to distance themselves from any
Possible association with the movement.

This first example ofa women’s movements was strategically moderate and
aPpeal ed to the universality of human nature. In Lu’s own words, it accepted the validity
0f marriage. the ethic foundation of the society, and the feminine nature of women by
setting the issue of freedom of sexuality aside to reduce possible objections. However,

the advancement of the women’s movement inevitably provoked traditional patriarchal

U1
[\J

attacks and harassment from the KMT. because the idea of feminism was clearly contrary
to the KMT’s women’s policy. The permissive women’s organizations under the KMT”s
corporatism advocated preserving the patriarchal traditions and roles. according to which
women should serve as good mothers. virtuous wives, and voluntary workers. Given the
conflicting directions of the women’s movement with the KMT’s policy. the movement
could not continue. With all the frustration and harassment from the KMT that followed.
Lu changed her life and went to study abroad. later becoming an opposition activist.
arrested and sentenced to 12 years in prison following the Kaohsiung Incident (discussed
later).
In 1982. to further the cause of feminism. a group of feminists formed the Women’s
New Awakening Association. which became the only association openly promoting
feminist and women’s rights under the KMT’s tight control. The New Awakening
worked to raise female consciousness and voice feminist opinions through the means of
publishing a monthly magazine. However. the New Awakening magazine was
considered radical and was not readily circulated in the beginning. In challenging the
traditional gender roles and culture, this association also sponsored annually thematic
activities regarding. and drawing attention to. women’s issues. and initiated discussions
on the injustice of women's legal status. They also carried their concerns into legislative
politics by seeking to change this legal status. Despite some practical difficulties
(management and finance). the organization succeeded in serving as the center of the
feminists’ movement. and coordinated the joint activities ofother women’s groups.
With the lifting of martial law in 1987, a variety of women’s organizations were
for med and women’s movements gained a new dynamism. In 1987. the New
Environmental Association was established to recruit housewives to advocate consumer
and enVironmental protection rights. In the same year. the Taiwanese Women’s Rescue
ASSOCiation was formed by some female lawyers and other concerned citizens to provide

legal Services for child prostitutes. Also the Warm Life Association was established for

supporting divorced women and widows. In early 1987, the Rainbow (a project for
rehabilitating escaped child prostitutes) and the Awakenings projects joined with other
human rights, religious. and women’s groups to demonstrate against the abuse of child
prostitutes on Hwa-hsi Street in Taipei. This was the first time that Taiwanese women
had organized a mass protest on the street.

The multiplicity and diversity of women’s organizations and activities demonstrated
that they were not only concerned with the traditional women’s issues ofchild and health
care, but also with broader social issues which might be contrary to the government’s
agenda. For example. they not only voiced their concerns regarding such matters as
prostitution, pornography, and discrimination in the workplace. but also participated in
protests related to consumption. environment and pollution. minorities, and the
aborigines’ land struggles (Clark and Clark 1997; Farris 1994). The diversity ofthe
activities and functions of women’s groups may not have necessarily furthered the cause
of women’s rights. but it did elevate women’s status through an increasing degree of
participation in social issues and public affairs.

Generally speaking. although the women’s movement existed before
democratization. its influence was limited. However. as democratization took place. the
Women’s movement embraced a period of transitional diversity. Facing the opposition of

a still largely patriarchal society and authoritarian regime. some of the women activists
persisted with the women’s movement and sought social reform as outlined above. On
the other hand, some of them joined the opposition movement. hoping to bring down the

KIVI—r's authoritarian regime.

Democratization
Democratic transition and women
The dynamics of democratization have changed the political landscapes and gender

pOIitics of the past decade. Under the authoritarian regime. most governments are

54

concerned with national security and stability and thereby prescribe their members certain
roles to achieve these goals. Women were mostly deemed to be the guardians of
traditional family values and social order. Given the history of rights granted from above.
the autonomous women’s organizations or movements found it hard to emerge into the
light. In Taiwan. under the authoritarian regime. women performed traditional roles
ratified by Confucian culture and the KMT’s ideology. As already pointed out, the
political roles played by women were extremely limited. Thus women who wanted to
hold public offices had to make their way either through local faction’s recruitment
systems or gender quotas. Furthermore. the KMT’s corporatism had atomized women
away from participating in any autonomous collective action. Even though women’s
movements emerged in the 19705. before the democratization proceeded. their influence
was not great.

However, Waylen (Waylen 1994) argued that some authoritarian regimes did allow
one section of the population, women, the space to mobilize, and that, in the absence of
conventional politics. they tended to allow women to develop new ways of practicing
politics. Comparing the differences between socialist and Latin American countries.
Waylen found that the latter gave women more space for political mobilization. In the
former societies. such as those in Central and Eastern Europe, the government. with a
symbolic commitment to gender equality. did include a considerable number of women in
politics. But a high proportion of women politicians were co-opted into the
establishment's politics and the women were. essentially. little more than tokens. On the
other hand, in Latin America. women were relegated to the protection of the traditional
values of God, family. and nation. Because the military government often did not see
Women’s activities as dangerous enough to warrant repression, this gave women the
Opportunity to mobilize their causes. For example. one of the first mass protests against
the Pinochet government in Chile was mobilized by women. who used their traditional

maternal roles to protest against the regime. That made it hard for the Pinochet regime to

1.11
U1

launch any attack since the women claimed to be fulfilling their maternal roles.

Like the situation in Latin America. women also played a significant role during the
transitional process in Taiwan (Clark and Clark 1997). In the 19705. social movements.
human rights groups, and political oppositions began to take root in Taiwan. In the
19805, these social and political opposition movements had been able to expand,
campaigning through a nationwide network of political alliances. A semi-opposition
group (Tang-wei. or non-KMT) was formed advocating a greater measure of
liberalization and democratization. The opposition movement was interrupted by the
Kaohsiung Incident of December 10. 1979. a large demonstration against the ruling party
which led to a violent confrontation with the regime. The regime responded by arresting
and jailing the major opposition leaders. After the event. to fill the evacuated leadership
positions following the arrests. the wives of the demonstration organizers became
symbolic leaders of the opposition. The absence of men in the opposition brought a new
wave of women into the opposition movement to continue the Tang-wei cause. Those
women without previous political experiences and involvement were motivated by the
KMT’s suppression and imprisonment of their husbands. and chose elections as the
channels through which to voice their dissent regarding political prosecution and the need
for liberalization. Running for office as relatives of political prisoners. and with the
voters’ sympathy and resentment of KMT injustice behind them. these women won a
stunning share of the vote in the 19805’ elections. These campaigns were run for moral
and symbolic reasons rather than in the hope of actually gaining political power.
However. by so successfully publicizing these moral concerns. these women were
advocating the opposition cause (Clark and Clark 1997).

The role women played at the beginning of the democratic transition was to create a
Space for which conventional politics does not allow. In Taiwan, the elected wives of the
arrested opposition dissidents continued the push toward democracy until 1986, when

they helped establish the Democratic Progressive Party (DPP). Later. as democratization

56

moved on, male leaders of the opposition were released from jail and reassumed control
of the DPP. As the politics returned to normal. most of these women. who had acted as
surrogates for their husbands. have not gone on to gain top leadership positions in the
DPP. Some of the surrogates withdrew or transformed their resources into family
politics. while others remained active. accumulating their own political careers based on
their past accomplishments.

Instead of being surrogates of absent male leaders or concomitants of political
families. some women played an independent and autonomous role in the opposition.
coming into politics in an incremental way. either working as assistants of opposition
politicians or joining the democratic protest movement. They had had first-hand
experience of being victims ofauthoritarian oppression. This type of woman, who had
grown under the authoritarian regime and had run for political offices successfully, was
relatively rare in the 805. They had been well aware of the reality of politics while in
office, but they still held rather objective attitudes toward their own goals and ideals
(Chiang 1996). Most of these independent women remain active in politics since their

political participation is as a result of their own long-held aspirations.3

The impact of democratization on women’s representation

As noted. no necessary connection exists between playing an important part in any
stage of the process of democratization and having a favored status during the period of
consolidation. In some cases. normal party politics pushed surrogated women back into
their subordinate political roles; in others. independent women politicians converted the
high profile they had achieved into institutional representation. It is important to evaluate
the impact of this reconstitution of democratic politics on women’s political participation

——‘

3 . . . H. . . . . . . . .

For example, whlle serv1ng as pub|1c officers. Chen. Clue 15 still enthus1astic for human rlghts; Lu, Hs1u-
Lien is still pursuing women’s independent status in the society and working for Taiwan’s international
Status,

on a general level. I examine the impact of democratization on women’s political
representation from the following two aspects: electoral and party political.

With respect to electoral politics. democratization requires open and regular
elections that provide women with the opportunity to vote and be elected. However. the
percentage of female representatives does not increase consistently as democratization
proceeds. If one compares statistics in the National Assembly before and after
democratization in the late 19805, the percentage of women representatives can be seen to
have decreased from 19.0 percent in 1986 to 12.9 percent in 1991. In the Legislative
Yuan election. the figure increased from 8.0 percent in 1985 to 13.9 percent in 1989. but
decreased again to 10.6 percent in 1992. In the mayoral and magistrate elections. which
control the most important executive resources. women won very few seats both before
and after democratization.

In terms of running as candidates. except in the national assembly, women’s degree
of participation has marginally increased. but the change has not been substantial.
However. the success rate of women candidates has actually decreased since
democratization began in 1986. In the Legislative Yuan. the success rate of women
candidates dropped from 66.7 percent in 1987 to 37 percent in 1991; in the National
Assembly. the rate decreased from 64 percent in 1986 to 56.8 percent in 1991(Chou
1995). Clearly. the percentage of women in the representative bodies did not rise
significantly as, even though more women ran for election. the success rate of women
candidates did not increase in line with expectations.

With respect to party politics. the open competition among parties - the KMT. the
DPP, and the NP - also creates new space for women to enter party politics.
Theoretically, because parties struggle for the majority’s support. they must mobilize
every possible political resource and appeal to different constituencies. As a result of the
perceived importance of women’s mobilization. parties may well try to gain women’s

Support by preposing more women candidates in the ballot.

58

.u.

In terms of the number of female candidates within each party after democratization.
except in the 1989 Legislative Yuan election. the percentage of the KMT female
candidates has decreased. whereas the relative number of DPP and NP female candidates
has increased. At the local level. the KMT’s faction domination has long prevented
independent women from running as KMT candidates. In terms of party nomination. all
three parties have increased their female nominees across time. Among the three parties.
except in the 1989 legislative election. the percentage of KMT female nominees has
always been lower than that of the DPP and the NP after 1986. Generally, therefore.
women have received higher levels of electoral nominations from the new parties. which
suggests that new parties tend to offer more opportunities for women. However, there
has been no significant increase in women’s electoral representation and appointment to

top government posts after democratization (Chou 1995).

The change of the Legislative Yuan

Significant change has also occurred in the legislative politics after democratization.
Before democratization. the majority of the members of the Legislative Yuan were the
tenured members elected in Mainland China. even though the supplementary members
have increasingly formed a much higher percentage of those who actively participate in
the legislative process.4 The elected supplementary seats as well as the tenured members
in the Legislative Yuan were subject to the KMT’s manipulation. and the Legislative
Yuan was viewed as a rubber stamp for decisions made. essentially. from the KMT party
center (the Central Standing Committee). Because the KMT’s nomination made it
extremely likely that the member would get elected (Ho 1986). most elected legislators
followed the KMT party line in the Legislature Yuan.

__n

’ Even the supplementary members constituted only a tiny minority in the Legislative Yuan, in 1983. they
Performed 52.7 percent ofthe articulation in the legislature. whereas the majority of the tenured members
never presented any ideas or spoke less than once per session in the legislature (Ho 1986).

59

The party dynamics in the legislative politics changed in the mid-19805. The DPP
(Democratic Progressive Party). the first opposition party. was formed in 1986 and
legalized in 1988. In I986. 13 DPP members were elected and entered the Legislative
Yuan. The number ofopposition members was too small to wield significant influence
on the legislative process. except to disrupt the normal operation of the legislature. The
1989 legislative election was the first nationwide election (the founding election)
following this legalization of the DPP. the first opposition party. and democratization was
thus institutionalized. In the 1989 supplementary election. the DPP won almost 30
percent of the vote, whereas the KMT earned 59 percent. With a critical number (21
seats), the DPP members were strong enough to effectively filibuster the KMT’s bills and
were able to attain a quorum in initiating bills and amendments. The vote share of the
KMT has steadily eroded. whereas that of the DPP has increased over time.

The legislative sessions from 1948-1992 were termed the first term of the
legislature. In 1992, the tenured cohorts involuntarily retired and the entire Yuan was
subject to election on a three-year basis, which was called the second term legislature.’
The party competition further intensified in 1993 as the New KMT Alliance registered as
the New Party (NP). The NP. with a significant voter base for a third party. gained few
seats but maintained strong vote solidarity and has often fulfilled a "balancing" role on
the floor and in committees. The DPP. combined with the NP, could form a majority of
the vote in the Legislative Yuan (Tien 1996). Faced with the DPP’s and the NP’s
challenges, the KMT’s dominance in the decision making process has been gradually
eroded, and the whole ecology of the Legislative Yuan changed. I show the changing

__

’ Since 1992. there have been four types of legislators in the Legislative Yuan. The first one is the
territorial representative elected by SNTV. The second one is the party representatives. proportional to the
vote shares party gain in territorial election.[( total votes for party A/total votes)* total PR representatives]
The threshold for the PR is 5% ofthe total vote share. The party representation replaces the system of
occupational group representation and gives the party more power selecting its representatives. Still the last
two types of legislators. which constitute a small number ofthe representatives. are 6 overseas Chinese
delegates elected by party proportion lists: and 6 minority representatives elected by aboriginal voters.

60

party composition of the Legislative Yuan over time in Table 2-4.

Table 2-4. Proportions of votes and seats of major parties in the Legislative Yuan
elections.l 986-1998

 

 

Year Total KMT % DPP% (tangwei) NP % Others
Seats Seats Votes Seats Votes Seats Votes Seats Votes
1986 73 80.8(59) 66.7 16.4(13)24.6 2.8(1)8.7
I989 101 71.3(72) 59.2 20.8(21) 29.9 --- 7.9(8) 10.9
1992 161 63.4(102) 52.7 32.9(53) 36.1 --- 3.7(6) 11.2

1995 164 51.8(85) 46.1 32.9(54) 33.2 12.8(21) 13.0 2.5(4) 7.7
1998 225 54.7(123) 46.4 31.1(70) 29.6 4.9(11) 7.1 9.3(21) 16.9

 

Notes: cell entries are the percentages each party received in elections. Figures in
parentheses indicate the number of seats ofeach party.

Sources: The data from the 1986 to 1995 elections are based on Sheng. 1996. p.13. The
data on the 1998 election are based on the web-site report of the Central Election
Commission. Taiwan.

In addition to party competition, the KMT and the DPP further split into several
opinion groups. The fragmentation has made it difficult for the party leaders to govern
the legislative process. In the past. the KMT leadership was able to impose party
discipline on its factional party members given the electoral advantages ofa KMT
nomination in the Legislative Yuan election. However. with democratization and
increasing electoral competition. the relationship between the KMT’s central leadership
and its legislative members has been weakened. Given the declining levels of KMT
discipline. the ruling majority party has lost its ability to manage the legislative process.
As a result of increasing party competition and loosened party discipline. cross-party
voting has also become very common. Bills were frequently held up in committee and on
the floor (Tien 1996). Put simply. the function and ecology of the Legislature Yuan has
been altered.

Furthermore. the legislative function has evolved to reflect the changing party
dynamics and new political reality following democratization. In the past, legislators

Were ‘used to paying most of their attention to interpellation as a gesture of administrative

61

(L

(L

oversight. However. the real policy impact of interpellation was limited. In the
watershed 1989 legislative term. most of the tenured legislators grew older and attended
fewer Yuan meetings. and the 101 newly elected supplementary members had gradually
taken over almost all the legislative power within the Yuan. Elected legislators not only
oversaw the executive in interpellation sessions and performed constituency service. but
also began to exercise policy initiatives and to participate actively in committees and
floor meetings. The executive department will still lead policy initiatives in many
instances. of course. but the policy steering power. including the informal process of
party bargaining and the formal process of agenda setting and policy debate. has
gradually transferred itself to the institutionalized Legislative Yuan.

Legislators have become more active and autonomous in the decision-making
process. Before the 19805. most bills were initiated by the Executive Yuan and
legislators just voted for them. In the late 19805, legislators started speaking more
frequently. In terms of speaking for bills. on average, a legislator spoke less than 20
times per session during the period 1973-1980, 32 times per session during 1981-1983.
and more than 70 times per session after the mid-19805. In terms of initiating bills.
before the 19905. the average number of legislators’ initiations was less than one per
session. However, from the beginning of 1991. legislators began sponsoring bills
aggressively. The average number of bills sponsored rose dramatically in 1993; there
were about 6.5 initiations per person at the break in session in the second term of the
Legislative Yuan (Sheng 1996). The Legislative Yuan. which used to serve as a rubber
stamp during the authoritarian era. has turned into a true law-making body and an
important policy forum.

The increasing levels of party competition and legislative autonomy have transferred
matters of policy decision from the centralized top-down KMT dominance to the
Competing "bottom-up" representative politics. Women began to protest the right to

demand government to meet their needs. to incorporate their concerns into policy

discourse. and to become political agents in bringing about changes. Parties have tried to
mobilize women’s support by placing women’s issues on electoral and legislative
agendas, and. there have been a number of policy initiatives aimed specifically at
women’s concerns.

A summary

Sociopolitical life in Taiwan is generally structured upon an informal kinship and
faction network. which precipitates a "patronage-led" electoral culture and discriminates
against the concept of female political representation because of their linkage to the
traditional patriarchal culture. Thus women must mobilize resources from supportive
natal and marital families to gain political power (Clark and Clark 1997). Nevertheless.
the introduction of local elections based on the system of SNTV. and the special
provision of the reserved—seats system for women. has countered many of the traditional
cultural disadvantages facing politically ambitious women and has created a beneficial
political opportunity structure for women to gain political power. These electoral
provisions create incentives for parties to nominate women candidates, which triggers
further election gains by women.

Rapid modernization has also removed the barriers against the entry of women into
public life. Increasing opportunities in the fields of education and employment have
improved women’s social mobility and level of political participation. Moreover, the
rising women’s movements and political transition have further impinged on the
patriarchal power structure of pre-democratic Taiwan. As a result of these favorable
changes in the sociopolitical environment. female entrance to representative politics no
longer relies on the approval of political families or factions or a reserved-seat system
administered by the state.

In general. this chapter has focused on women’s political status in Taiwan’s
changing sociopolitical context. This lays the foundation for understanding the gender

dynamics visible in the legislative politics to which I will now turn.

CHAPTER 3
THEORETICAL HYPOTHESES AND RESEARCH DESIGN

Theoretical hypotheses

Liberal theorists assert that. regardless of their social and cultural backgrounds. and
thus without making any distinction based on identity differences. all people should be
treated equally. Therefore. representatives can represent anything or anyone without
sharing particular identity features and. consequently. a random relationship is expected
between identity and representation since identity characteristics will not influence a
representative's legislative performance. On the other hand. the politics of difference
posits that identity characteristics do matter and that they shape the policy interests of
deputies. Therefore. this approach implies that gender distinguishes male from female
legislators with respect to the substance of policy representation. However. it is clearly
the case that these two theories contradict each other and. in order to resolve this issue. I
propose to re-examine the empirical base ofthis theoretical debate. focusing primarily
upon the representational politics of Taiwan.

A study of women in Taiwanese legislative politics in the 19805 found that those
who entered politics through the reserved-seats system differed little from their male
counterparts in terms of role orientations. types of legislative activities. and legislative
effectiveness. The only exception to this. as already observed. was that they were less
active than man as regards actually sponsoring legislation (Chou. Clark. and Clark 1990).
However. despite this apparent lack ofa clear difference between the sexes. certain
differences can now be seen to have been underestimated. One reason for this was the
urgent concern at that time to demonstrate that a female could compete with men on the

same terms. rather than to identify different policy interests between the genders. In other

64

words, attempts to demonstrate that there was a difference between the sexes were
regarded as undesirable in the 19805. Consequently. the concerns and aims of the
researchers themselves have. to a certain extent. attempted to reflect political reality. by
putting more emphasis upon comparing male and female equality as regards opportunities
and capabilities. but neglecting to examine the meaning of female representation to
women and the differences that gender might make.

Generally speaking. the democratization of Taiwan since 1987 has given rise to
different voices in a more equal tone. In the 19905. a heated debate upon identity politics.
such as ethnicity (the mainlanders versus the native Taiwanese). ideology (independence
versus unification). and gender. have combined to foster a new awareness of the “politics
of difference” that such multiple identities can generate. However. to date, the majority
of studies have identified ethnic identity and conflict as the most important factors in
understanding party coalition. ideological confrontation. and social fissures in Taiwan.
As a result of this focus, the possible relationships between gender identity and group
representation have. as already suggested. been largely ignored. However. as more issues
emerge that span the entire political spectrum. and competing political groupings succeed
in inspiring very different sections of Taiwanese society to take an interest in them (Chu
and Lin 1996). it is my belief that matters of gender will become increasingly important.

As already noted. the importance of the role of women in Taiwan has grown
significantly in a number of ways. Socially and economically. due to increasing
educational and employment opportunities. and the effects of social modernization,
women have been able to compete on more equal terms with men in more occupations
and activities than was previously the case. In electoral politics. since the late 19805.
women have been able to exceed their minimum quotas by a significant margin. The
intense and growing competition between the main political parties - the KMT. the DPP.
and the NP - has forced these organizations to take note of the new and growing

electorate and to respond by putting forward more female candidates. There is also

65

evidence to suggest that women have ceased to be mere “tokens” in the political process
— an idea generated by a minimum quota. Rather. they have become active and effective
campaigners - a political force in their own right.

Moreover. the flourishing of various women's movements since the late 19805 has
pushed women’s issues to the foreground of various political agendas. with debates for
gender quotas sometimes altering these agendas significantly. In addition. following the
murder of Peng in December 1996. a women’s caucus was formed within the Legislative
Yuan. The Coalition for Promoting Women’s Rights. initiated by six DPP women
legislators and endorsed by female legislators from right across the political spectrum. is
one of a handful of factions that have been formed across party lines in the Legislative
Yuan. Considered together in this manner. it becomes clear that a number of factors have
combined to create a favorable environment for women’s interests to emerge and for
differences of gender to become increasingly important in representative politics.

The argument of identity difference posits that women and men are different.
whether for biological. sociological. or psychological reasons. and that these differences
will inevitably affect representative politics. If it is assumed that female legislators are
more likely to be concerned with “women’s issues” than their male counterparts. then. to
accurately gauge the impact of gender on legislative politics it becomes necessary to
specify just what these “women’s issues” consist of. Therefore. I will begin this section
of the chapter with an attempt to identify the policy domains in which gender differences
may have an effect. Moreover. as I discussed in the theoretical section. different kinds of
representation exist; namely. descriptive. symbolic. and substantive representation. To
examine how these different modes of representation are actually employed in Taiwan. it
is necessary to distinguish them. Consequently, in the second section 1 will elaborate
upon these different forms of representation. and then put forward a series of sub-
hypotheses along the lines of the dimensions identified. In summary then, the following

two sections will address the notions of (1) the policy areas and. (2) the legislative

66

dimensions. within which female and male representation would appear to differ.

The policy domains of gender difference

I assume that gender differences in representative politics are likely to be most
apparent in policy areas that concern themselves with “women’s issues”. In other words.
female legislators are more likely to devote their time and effort to women’s issues.
However. to better assess the impact of gender in legislative politics. I must first specify
what I mean by “women’s issues”. Little agreement exists on whether significant gender
differences are present or not in politics and what such differences might involve. Some
feminists have argued against the theories of gender difference. and the existence of an
objective set of women’s interests. for the reason that it might “freeze” what are multiple
and shifting identities and instead label women as a single and indivisible grouping.
Arguing against any such all-embracing sense of identity based upon gender distinction
alone, these post-modem theorists instead suggest that heterogeneous difference coexist
(Rhode 1990). Therefore, embroiled within these theoretical disputes. an objective
definition of “women’s interests” seems impossible.

Even though some feminists deny the essential. foundational. and biologically
determined differences between the sexes. the majority of feminists consider it dangerous
to completely deny the existence of gender difference. Notwithstanding the dispute over
the definition and content of an objective women’s interest. most feminist theorists still
believe that gender is a fundamental organizing principle of social life and of human
consciousness. and the majority share some basic beliefs as to what constitutes women's
interests (Jonasdottir 1988). As Sapiro (Sapiro 1991) put it. gender schemes are
important in structuring people’s perception and behavior. “Women’s shared gender
gives them shared political interests. Women face certain problems that need to be
addressed through politics and. for the solving of those problems, their shared gender

takes primacy over other types of identities and characteristics that may divide

67

them.”(p. 189)

In light of the conflicting nature of theories of gender difference. it is necessary for
me to define exactly what I mean by gender difference in politics and then clarify which
values or priorities will form the basis for a possible transformation of policies. I attempt
to apply a broader definition of gender difference in policy domains here. and the
women’s issues referred to herein encompass not only the concerns of different feminists,
but also the interests that originate from female physiological experiences. The former is
subjective in the sense that women’s interests are here equivalent to what women want.
These can be ascertained via an examination of the public policy preferences of women’s
groups. and I will refer to these desires as “feminist issues”. On the other hand,
according to Abby Patterson, women's issues could also be defined objectively through an
examination of the value systems of female culture. Gender-differentiated interests. as
they pertain to women and men. refer to the fact that the sexes participate in different
activities. work with different things. have different responsibilities. and involve other
people in different ways. The differences in interests originate from men and women
having undergone different experiences due to being from different social spheres:
namely, the public as opposed to the private (production versus reproduction). These
distinct physiological experiences also assign men and women different social agendas.
Therefore. by way of providing an operational definition of “women’s issues”. I specify
two main types of women’s issues as below:

1. Feminist issues - defined by some basic principles or policy goals and agreed
upon by most feminists. These include removeing ascriptive criteria in the allocation of
social resources by eliminating discriminatory sex roles, broadening the choices open to
women by removing social and legal constraints. addressing women's needs from
different psychological or socioeconomic perspectives. and generally improving women’s
status in society (Carroll 1984; Saint-Germain 1989; Tamerius 1993).

2. Caring and communication issues - matters traditionally associated with the

68

domestic concerns of women. such as children. education." health care. and social welfare
(Darcy. Welch. and Clark 1994; Saint-Germain 1989; Sapiro 1981; Thomas 1991). As
suggested above. most feminists agree that the division of production/reproduction
ascribes different social roles and functions to men and women. Traditionally men and
women are separated into the public/private spheres. Furthermore. women’s continued
responsibilities as the primary givers of care within the family unit assign women
different characteristics and interests from men. As a result. the processes of sex-based
socialization. and division of labor, have situational influences on gender behavior.

By contrast, on the basis of the public/private dichotomy, such policy areas as
economic development. and government organization. have traditionally been familiar to
most men. and are thereby regarded as “men’s issues”. Therefore. it is equally expected
that men will invest most of their attention in these policy areas - dealing with such
matters as governmental roles and rules. political reform and competition. and economic

resources and their allocation.

The legislative dimensions of gender difference

The politics of identity difference begin with the general assumption that group
identity (gender) does indeed matter and that it shapes the form and content of
representation. However. the ways in which females and males actually differ in terms of
representational politics remain to be explored.7 Given the complicated nature of the
legislative process. it is imperative to first clarify the various dimensions of
representation that exist and, therefore. I have positioned my inquiries upon this point

within the theoretical framework of the three types of representation - namely.

 

" In Taiwan. education. culture. and communication are conventionally linked together within the same
category of education. which deals with interpersonal relationship and is the profession predominated by
women.

7 Many past studies have tested the gender difference in the legislative politics in scattered research without
a systematic examination ofthe nature and dimensionality of representation as a whole. and the results as
we have seen in the literature review are also inconclusive.

69

descriptive. symbolical. or substantive (Pitkin 1967).

I assume that for substantive representation to be successfully sustained, female
representatives might be both attitudinally and behaviorally different from men with
regard to the presentation of female concerns. For symbolic representation to prevail.
women representatives might be attitudinally linked to women’s concerns. but not
necessarily perform accordingly. In other words. female representatives will stand for a
particular constituency and their concems. but not necessarily act upon their subjective
beliefs. F inally. for female legislators to represent women in a merely descriptive way,
they need only share the gender characteristics of women in general. and need not be
identified with them. or act for them. Beneath these modes of representation lie two very
different dimensions - the behavioral and the attitudinal. Even though the way certain
female legislators represent might not easily fit into either category. this framework is
extremely useful in helping to locate and investigate the different types of women in

Taiwanese politics.

Behavioral hypotheses

The study of legislative behavior must take into account the multiple-layer nature
and different stages of the legislative process (Hall 1996). Hall argued that the
representatives engage in various legislative activities other than voting on bills in order
to pursue their concerns. The possible types of legislative participation range from
agenda setting. offering amendments, and speaking during mark-up. to voting and the
mere act of attending. Many feminists have also questioned the measures used by
traditional behavioral researches - for example. studying voting patterns alone might bias
the result we would obtain when attempting to identify gender differences (Burrell I996:
Tamerius 1993).

Therefore, it is vital to understand Taiwan's legislative structures and processes

before attempting to study legislators’ behavior. Each legislative term in Taiwan lasts for

70

three years and the Legislative Yuan holds two sessions per year. With respect to
structure. the Legislative Yuan uses a committee system. and there are a total of ten
standing committees covering economic. finance. budget. transportation. interior and civil
affairs, government organization and administrative affairs. education and culture, legal
and judicial affairs. foreign affairs. and defensive matters. The committees are based
upon the labor division of the executive departments and are each responsible for bill-
examination within their own subject areas. Each legislator is required to register to join
one committee at the beginning of each session and most legislators will sit on a number
of different committees during the three-year term. Amongst the committees, some are
inevitably more popular than others8 and, whenever demand for a popular committee
exceeds the maximum quota (18). legislators are required to cast lots. Usually. however.
the parties will have attempted to coordinate committee membership long before this
stage has been reached. In addition. members of each committee elect three Chairpersons
at the beginning of each session and these Chairpersons take turns to preside over the
workings of the committee and setting agendas (such as which bills will be discussed
each week).

With respect to the legislative process. it involves both formal and informal
dimensions. However. as recorded data upon such informal activities as bargaining and
log rolling is not available. I am forced to focus upon the formal dimensions. The formal
process of legislation from initiation to passing ofbills involves three reading stages. The
first reading is to report the bill-initiation (both govemment- and legislator-sponsored
bills) to the entire Yuan and then to refer the bill to the relevant committee for discussion.

examination, and mark-up. After the committee examination is complete. the committee

 

” The most prestigious committees are the economic. transportation. and interior committees. The
transportation committee deals with issues such as transportation construction. highway. airport. traffic.
etc. Most ofthe issues within the transportation jurisdiction are key to local development. therefore most
legislators favor this committee for their constituency interests. The interior committee deals with issues of
political reform. elections. police. local development. labor. and civil affairs. Because of the
comprehensiveness and influence of the interior committee. it is also one of legislators’ favorite choices.

71

report is then sent to the Yuan for a second reading. during which the legislators can
usually be expected to commence with a general discussion and then move on to an
article by article review of the bill. Once further conclusions have been reached, another
Yuan meeting is arranged (sometimes immediately following the second reading
meeting) for the third reading. in which the wording of the bill is examined and. finally.
voting takes place. Generally speaking. the three reading procedures are for the purposes
of agenda setting and policy formation. Once bills have been initiated. in the first
reading, and sent to the committees. the majority of committee and Yuan meetings
become forums for intensive discussion of the bills.

To embrace the different aspects mentioned above. both structural and as regards
legislative processes. this study observes legislative participation from two perspectives -
namely. a horizontal division of labor (structure) and the vertical legislative dynamics
(processes). Therefore. hypothesis one focuses on the horizontal division of labor in the
legislature by examining the choices the deputies make regarding which committee to
join, whereas hypothesis two focuses on the deputies’ participation in the dynamic
legislative process from agenda setting to policy formation. Consequently. the general
behavioral hypothesis that female legislators are more likely to engage actively in
women’s issues than are their male counterparts is examined on both a structural and

dynamic level.

Hypothesis 1: In terms of horizontal institutional participation. female representatives

 

are more likely than their male counterparts to work within committees dealing with
“women’s issues” defined earlier.

Committee assignments set the primary structural division of labor within the
representative body. and provide institutional power for legislators over the policy-
making process. Studies have found that women representatives tend to choose the care—

oriented committees rather than the technological. economic. or other types of

committees. Data from the Inter-Parliamentary Union (IPU) regarding certain countries -
namely, Austria. Denmark. Germany. the Netherlands. Norway. Spain, and Sweden -
demonstrates that women are most numerous in the areas of family. social affairs. health.
and education (IPU) 1992). Comparing the participation of women in legislative
committees in thirteen Westem countries. only in Sweden did the extent of female
participation in the defense committee approach the extent of their representation in the
parliament as a whole. In all other countries. female representatives continue to register
or be channeled into committees dealing with health and welfare.

Two possible explanations account for gender differences as regards committee
assignments. Firstly. some claim that the difference is due to a type of institutional bias
or discrimination. beneath which women tend to cling to the less prestigious committees
and find themselves closely associated with these committees. Alternatively. some claim
that the differences can be explained by gender socialization. e.g.. women possess
different training specializations. make different professional selections. or maintain
different policy priorities (Kelly. M.A.. and Horn 1991).

In Taiwan. studies in the 19805 showed that women were more likely to serve on
both the education and interior committees (social issues and civil rights) in the
Legislative Yuan. The OLS regression showed that. when other explanatory factors were
taken into account. women were 32 percent more likely than men to have served on the
education committee and 18 percent more likely to have sat on the interior committee.
dealing with most domestic policies. However, gender made no statistically significant
difference in any other areas - women did not appear to be under-represented. except in so
far as more men than women had been members of committees related to economic
affairs. Generally speaking. women did not suffer bias in committee selections. It was
found that party affiliations and ideology had no effect in committee assignments. while
the importance of experience accounted for the fact that seniors were more likely to serve

on prestigious committees (Chou. C lark. and Clark 1990). In this study. I hypothesize

73

that gender division in terms ofcommittee assignments does exist. however, I will
suggest that it is due to the specific needs that arise from the different socialization and
psychological conditions and interests of women, rather than any form of gender
discrimination. Therefore, I will examine whether women are more likely to join
committees such as the interior and educational committees than men and the possible

explanations for the pattern.

Hypothesis 2: In terms of actual legislative participation, female representatives are more
likely than their male counterparts to engage in legislative activities in the area of
women’s issues. from agenda setting to policy deliberation.

In light of the complicated legislative processes and multiple dimensions of
legislative participation in Taiwan, I generate four measurements for evaluating gender
difference in the legislative process. The level of behavioral legislative participation is
generally divided into two parts: agenda setting and legislation formation. Agenda
setting in the three-reading process involves two stages. as does policy formation. For
the former, the stages are that of bill initiation in the Yuan and chairing the committee
meetings. for the latter, speaking in the committees and speaking in the Yuan meetings.
Overall, then, bill initiation is clearly the first step in getting an issue heard and,
consequently, addressed. However, it is to be noted that many bills are proposed in each
session, and that decisions as to which bills are discussed and at what pace are the
preserve of the committee chairperson. Consequently. the speeches made in committee
and Yuan meetings represent potentially critical stages in the legislative process as well
as in the process of policy formation. directly affecting the formation and content of the
bills.

Therefore, dealing with these four” forms of legislative activity together. I

 

” Some legislative studies also include co-sponsorship and voting as indicators of legislative participation in
policy-making. However. two problems occur in using co-sponsorship for this analysis: first. co-

74

hypothesize that:

a. Female representatives are more likely than their male counterparts to initiate bills
related to women’s issues.

b. Female representatives are more likely than their male counterparts to chair
committee meetings examining bills related to women’s issues.

c. Female representatives are more likely than their male counterparts to speak in
committee meetings examining bills related to women’s issues.

d. Female representatives are more likely than their male counterparts to take part in

floor deliberation upon issues related to women’s interests.

Attitudinal hypotheses

The attitudinal hypotheses are formulated to explore the possible differences of
perception toward the idea of representation between the genders. Perception differences
build up the foundation for behavioral difference and. therefore. gender differences in
attitude are a necessary, but not sufficient. condition for gender differences in behavior.

This is especially the case in an arena such as politics. in which women have traditionally

 

sponsorship does not have a substantial meaning in terms oftelling us the individual legislator’s interests in
law making. Generally. representatives simply sign their names on others’ bill in seeking the exchange of
the support from others for their own bills. Some deputies have no idea what they have signed. in fact. on
some occasions. they even endorse contradictory versions of the same bill. Since the act of co-sponsorship
does not require deputies to take the time in making decisions or collecting information. most legislators
co-sponsor their colleagues’ bills for various concerns (parties. districts, factions. committees, friends). The
co-sponsorship does not necessarily warrant the deputies’ real interests in the bills they sign. Second. no
systematic data is available on the co-sponsorship. Since 1995. all sponsorships are reported as only the
title and the main authors to the Yuan meetings. and are then referred to the committees. The originally
formal function of reporting all the co-sponsors ofthe bills to the Yuan meeting in the first reading has
been replaced by the informal functioning ofthe scheduling committee since then. Thus. we can no longer
find any systematic official records about the co-sponsorship. Likewise. the problem of applying voting as
a measurement is also related to the problem of substantial meaning and data availability of voting. The
problem of the significance of voting has been discussed earlier about how voting data might bias the
measurement. The problem of data availability is because many bills did not involve the voting process.
Voting tends to happen only for those important and controversial bills and articles. For example. the
voting rate in the political and social welfare bills are higher than those of other categories are (Sheng
1996). Considering my attempt to examine the overall differences. these two conventional measurements
are not included.

75

been prevented from active participation. Consequently, though female deputies could be
attitudinally associated with women’s concerns, they may be unable to carry their identity
characteristics into actual representative acts. Therefore, the symbolic representation of
female concerns is probably more common than substantive representation. Whether this
actually turns out to be the case or not. examining the perceptual differences between the
genders will clearly help us to better understand the calculations. constraints. and
motivations behind legislators’ decision making.

To evaluate whether gender makes a difference in terms of legislator's attitudes
toward representation. two questions are essential - whom do deputies perceive to be their
primary concern; and what do they consider it important to represent? The former
question relates to individual deputies’ level of constituency identification and the latter
to their policy priorities. and they are addressed in hypotheses three and four.

respectively.

Hypothesis 3: Gender differences are present regarding the perception of the
“representation of women”.

I assume that female legislators, because of their group affiliation and identity. and
whether it arises from sociological or psychological conditions. are more likely to
identify with female constituencies. and are more likely to have a personal commitment
to representing women as a group. Empirical studies have also shown that female
legislators are more likely than their male colleagues to see women as an identifiable
component of their constituency (Freeman and Richardson 1995; Reingold 1992; Thomas
1992) (Dodson et a1 1995). Therefore. I hypothesize that:

a. Female legislators are more likely than their male counterparts to think of
themselves as representatives of women. and to consider women as an important
constituency group with particular political concerns.

b. Female legislators are more likely than their male counterparts to believe that.

76

because of their gender. they are uniquely qualified for. or responsible ”for.

representing women’s concerns.

Hypothesis 4: Gender differences are present in the policy preferences and priorities.

In hypothesis two. I assumed that gender differences existed in actual legislative
participation in two stages, agenda setting and legislation deliberation. and I further
divided them into four measurements: initiation, chairship, speech in committee, and
Yuan meetings. A full investigation into every aspect of legislative activity is required by
this hypothesis. In addition. I agree with Tamerius (1993) that the most significant
difference between female and male legislators lies less in their willingness to support
feminist politics than in their willingness to initiate and guide policy processes pertaining
to those issues. To better assess willingness to lead policy process in such areas. I
therefore assume that female and male legislators have different policy priorities.

Because of gender difference in terms of socialization and life experience. women
and men tend to encounter different problems, associate with different groups, acquire
different policy expertise. and nurture different policy interests. As already noted.
empirical studies have shown that female representatives tend to give a higher priority
than their male colleagues to matters of women. children, family, and welfare (Boles
1991; Dodson 1991; Flammang 1985; Havens and Healy. 1991; Kathlene 1989; Leader
1977; Saint-Germain 1989; Tamerius 1993; Thomas 1994). Consequently. my
hypotheses are:

a. That women tend to place a higher priority on issues related to female issues than do
men.

b. That the legislative agenda might be changed if more women entered legislative
politics.
It is also expected that such perceptual differences regarding policy preferences and

priorities would markedly affect legislators’ agenda setting. policy choices. and

77

legislative.

Hypothesis 5 .' A strong and positive link exists between a female legislator's perception of
representing women and the policy preferences she will exhibit regarding women’s
issues.

I assume a positive link exists between subjectively identifying with women as a
constituency and granting women's issues a higher policy priority. If hypothesis three
holds true, then it should have strong implications on legislative priority in hypothesis
four, since legislators who view their constituency differently may act upon their

subjective perceptions rather than an objective definition of that constituency.

Research methodology

Two research methods

In the light of these two main forms of gender difference (behavioral and
attitudinal), I attempt to test these five hypotheses. and thereby evaluate the gender
dynamics present, in a Taiwanese context. To better discern the affect of the institution
in question upon female representation, it would in general be preferable to be able to
conduct either a cross—institutional or a longitudinal study, and thereby observe the
possible patterns of gender difference in a variety ofsituations. Some cross-institutional
studies of gender difference have been carried out in the United States - comparing two or
more states to “factor out” the institutional effect (Reingold 1992; Thomas 1994).
However. such an investigation across different legislative institutions is less desirable in
Taiwan. This is because. despite the fact that there are several different levels of
legislative body from the national to the local. only in the Legislative Yuan does one
observe the sort of complete legislative process that allows legislators to actually make
laws and be fully involved in policy-making. In other bodies. representatives are most

likely to exercise an “oversight of the executive” role rather than actually make policy.

78

making it difficult to perform a comparative study of them. Hence I decide to concentrate
on the Legislative Yuan. As regards a longitudinal study of both attitudinal and
behavioral dimensions. this would be almost impossible. if undoubtedly desirable.
Representatives are. after all. replaced every three years. and it thus becomes hard to trace
their attitudes toward public policies and perceptions of representation because they have
vacated their positions. The most I can learn about legislators’ attitudes is to elicit the
views and perceptions of those who are currently serving on the representative bodies.
My research is, therefore. divided into two parts - a longitudinal study of legislative
behavior and a series of interviews with legislators that attempt to gauge their perceptions
of their representational roles. Therefore. while Chapter 4 focuses on the longitudinal
study of legislative behavior from 1990 to 1997. and is set up to test hypotheses one and

two; Chapter 5 focuses on the interviews and tests hypotheses three to five.

Data collection

Data collection is accordingly divided into two main parts: a primary content
analysis of legislative activities to test the affect of gender on policy processes
(hypotheses 1 and 2). and face-to-face interviews with sample legislators to gain an
understanding of their perceptions of representation (hypotheses 3 toS).

Before actually collecting data. I first tried to gain some background information
from official documents, newspaper reports. and relevant records upon my sample
legislators in order to set up a primary data bank of demographic information. This
included such information as age. education. ethnicity, party. seniority, committee
assignment, institutional position. and the nature of their constituency (either urban or
rural). I set up this demographic data bank to facilitate further multivariate analysis. as.
with this supplemental information, one is better able to test whether gender is a
significant factor after controlling for other variables. just as hypothesized. After

obtaining this demographic data. interviews were conducted to test my attitudinal

79

hypotheses and to enable me to gain a more contextualized understanding ofthe

operation of the legislature.'0 The necessary data collection for the longitudinal study of

legislative behavior was then undertaken. after the second half of the interviews. And a

primary content analysis of the legislators’ activities was then sorted and coded. The data

sources for this study can thus be summarized as:

(1)

(3)

(4)

The Electoral Bulletin of Central Election Commission and legislators’ rosters.
From this data set. I obtained basic background information upon the legislators.
such as their age. education. party affiliation, seniority. occupational background.
and district information. especially the degree of urbanization evident therein.

The monthly legislation reports. These reports cover bills proposed every month by
first author. bills under committee discussion, and bills proceeding through the

Yuan meetings. The Secretariat Chief of the Legislative Yuan provided me with the
information and assured me that the names on the list were indeed the first authors
of the bills. I recorded bill initiation using this data.

The index system of legislative participation. This system provided information
concerning the Chairpersons in each legislative committee meeting. as well as the
frequency of the speeches the legislators made in the committee and Yuan floor
meetings. In the index record. each bill examined in committee is assigned one or
two subject areas that serve as the basis of bill classification.

The Legislative Gazette. The Gazette reports all the legislative activities proceeding
in the committee and Yuan meetings. including sponsorship. interpellation, election.
and voting. This serves as a reference for checking the content of the proposal and

draft versions of bills.

 

'0 During the interview process. some of my original behavioral measures (co-sponsorship and voting) have
been found to be less relevant and unavailable. yet the information obtained from the interview adds
another measurement (chairship) to my original measurements of the legislative participation. Therefore
the procedure of my data collection has been revised to accommodate the new measures of legislative
participation found during the interview process.

80

(5) The interview data. I interviewed a sample of legislators to explore and compare
male and female representatives’ conceptualizations and attitudes to representation.
The interviews took place between September 1997 and January 1998. during the

fourth session of the third term Legislature.

Method 1: Longitudinal study of legislators’ behaviors (i.e., hypotheses one and two)

Time span and unit of analysis

 

This longitudinal behavioral study is designed to test the first two sets of hypotheses.
The first hypothesis suggests that a horizontal labor division exists, along gender lines. as
regards committee choices and chairships. The second hypothesis states that different
degrees of participation are present, by policy subject, across different legislative stages.
The examination of these two hypotheses provide a relatively complete picture of the
legislative structure and processes by organizing into different aspects of legislative
activities, both static institutional arrangements and the dynamic legislative process.

I studied the legislators’ behavioral participation from 1990 to 1997. including the
eighty-fifth to ninetieth sessions of the first. second, and third term Legislative Yuan.ll I
had two basic reasons to focus upon this particular time span. F irstly, the 1990 legislative
session was the first legislative session held following the democratic elections of 1989.
Since 1987, the sociopolitical climate has changed, martial law has been lifted. the first
opposition party. the DPP. was formed, and the functions of the Legislative Yuan have
been revived. This is all in sharp contrast to the situation prior to 1988, in which the
Legislative Yuan could not be seen as an independent legislative body, but rather as a
rubber stamp for the KMT party-cum-administration. Under the KMT one—party regime.

most of the policies and bills were made within the party itself or the Executive Yuan

 

” I cover the 85th to 90lh session of the I“ temi. from February 1990 to January 1993; the 2"d term. from
February 1993 to January 1996; and the 3rd runs from February 1996 till the last session of my interview.
Jan 1998.

81

(Lee 1989). Therefore, since representatives did not play any significant role in law-
making. no meaningful data was available for study before 1988.

Secondly. prior to 1990. the majority of the representatives were senior members
who had been elected before 1947 from Mainland China. and had thus been able to
remain in their seats without the need for re-election. Though Taiwan had gradually.
from 1969. introduced new members into the Legislative Yuan through the
implementation of a system of supplementary seats. it was only following the 1990
supplementary legislative that these supplementary members started to play a significant
part in law-making. As Sheng noted, there is a large difference between the figures
concerning bill-initiation and speeches before 1990, and those that followed (Sheng
1996). Finally. my concentration upon this period provides us with a better point of
departure with which to compare this study to Chou. Clark. and C lark's pre-l990 study.
carried out under the authoritarian regime.

The unit of analysis is each legislator’s level of participation in each policy area. I
cover individual legislator’s participation in initiating. chairing committees, and speeches
in committee and the Yuan across three terms. In the first term. there were five different
types of legislators: the first (and making up the majority). being the senior
representatives who had held their positions since 1947. and the second being overseas
Chinese representatives. Voters did not elect these types of representative and, therefore.
they were immune from the need to represent the public and thus seldom participated in
the legislative process. The third type were the district representatives, with a total of 78
members. and the fourth type were the occupational group representatives. with a total of
19 members elected by the voters of five occupational groups. The last type was the
aboriginal representative, of which there were a total of4 members elected by aboriginal
voters. Excepting those senior representatives and overseas representatives who were
either not subject to re-election or were not elected at all, all the legislators were included

in my analysis. Come the second and third terms. all the senior mainlander members had

82

retired and had been replaced with newly elected members. and the group representatives

and overseas representatives were replaced by the party-list proportional representatives.

My sample includes all the elected members and party-list members. excepting those

members who had died or served for less than one year. and the President and Vice-

President of the Legislative Yuan. The President and Vice-President seldom participate

in the legislative process as regular members - instead they focus on generally facilitating

the proceedings of the meetings.

Among all the members included in my analysis. the gender and party distribution

is shown in Table 3-1.

Table 3-1 The sample legislators for the behavioral study

 

 

 

 

 

 

1St term 2"" term 3"" term
( 85-90 sessions)
Gender
Female 12 17 22
Male 86 138 137
Types of members
District 75 114 120
Group (PR rep.) 19 (group) 35 (PR party-list) 34 ( PR party-list)
Ethnic minority 4 6 6
Party
KMT 72 94 81
DPP 19 49 55
NP 0 8 21
IND 7 4 2
Total 98 155 159

 

 

 

 

Data coding and analysis of legislative participation (i.e.. the methods used to test

hypothesis two)

Hypothesis two involves a rather complicated series of measurements and

comparisons regarding legislators’ overall behavior. Because of this level of complexity.

I make especial efforts to clarify the methods of coding and analysis used to test it. I

haVe divided the legislative participation into four stages: initiation. chairship. speech in

 

the committees, and speech on the floor. I then examine the gender differences at each
stage and at two levels: aggregate and individual. Firstly. at the aggregate level, I want to
observe the group (and gender) differences in the levels of participation by treating
women and men as two different groups. I expect to reveal from this information
concerning how the two gender groups expend their time and energy upon different
policy areas at each legislative stage. Secondly, once we see such a pattern emerge. I
then pinpoint the specific policy subjects related to women’s interests and apply OLS
regression analyses to examine the gender affect after controlling for other factors.

In each measurement of legislative participation, I first classify bills into several
policy categories and then compare the overall differences between women and men. by

policy subject. as regards their levels of legislative participation.

1. Bill classification

At each legislative stage, I am particularly interested in the different policy domains
in which various deputies may take an interest. In order to pursue this, it is necessary to
classify bills by policy subjects of interest to the representatives. As the first stage of
specifying these policy subjects, I attempt to establish a general policy scheme to cover
all the bills proposed and discussed. Given my interest in testing gender differences as
regards policy interest, I first refer to the gender-based policy schemes of Thomas (1994)
and Saint-Germain (1989). and I adopt five categories from their policy schemes -
education, children, welfare. health. and feminist issues. Besides these five categories of
theoretical interest to me. each bill examined in the committees is assigned one or two
subject types in the Yuan index record. After browsing through all the bills’ subjects and
incorporating my theoretical interest. I established 19 policy-coding categories (see
Appendix 1).

- In coding the bills, those 19 categories might be expected to be completely

exhaustive and mutually exclusive. However, many bills do touch on two or more

84

subject areas at the same time and are thus referred to more than one committee for
examination. For example. the problem of child prostitution is an issue that embraces
both children and women (pornography). the provision of education for the retarded is an
issue of both educational and handicapped welfare, and subsidy for senior farmers is an
issue that concerns both agricultural policy and welfare policy. Even in a committee
meeting for the same bill. legislators may raise issues. or amend the bill, based on
different thematic concerns. Such multiple themes contained in one bill caused many
difficulties as regards their coding. In an effort to reflect the legislators’ own
classifications. I chose to base my decisions on the most prominent dimension of each bill
as presented to the public. Though I feel this to have been an effective approach to the
problem, I must. however. admit to my subjective bias in perceiving public perceptions.

In the second stage. a further collapsing of the bill categories was required in order
to show the pattern of legislators’ policy concerns as a whole. and to accommodate a
wide range of policy subjects. Therefore, in an attempt to further collapse the bills into a
more “macro-structured” scheme. I considered several schemes as references here, such
as Sheng’s scheme (1996), for studying the factors influencing legislative participation in
Taiwan, Chou. Clark. and Clark’s scheme (1990). for studying female representatives in
Taiwan, and Thomas (1994) and Saint-Germain’s (1989) gender-specific and theory-
oriented policy scheme in order to discern possible gender differences (see Appendixl for
a brief discussion of each scheme). In light of the two general policy schemes applied in
Taiwan’s case, and the other gender-based schemes. 1 reduced the 19 categories into
seven further categories.

Among the 19 categories. based on my theoretical hypotheses and some past studies.
I establish the traditional “women's interest” area by collapsing the categories of
education, children, health, welfare, and communication, together. I combine these
categories into a new category entitled “care and communication”. Another issue area

that I theoretically hypothesize to be gender specific is women’s rights (feminist) issues.

85

As to the remaining categories. I collapsed those categories with the most similar
characteristics into a single category. For example, I combined finance. budget.
economics. and local development into a category named “economics and development”.
political and government organization into one. judiciary and social order into one. and
defense and foreign affairs into one. The remaining categories included labor. farmer.
fishermen, ethnic minority. and the environment. Because these categories share
common characteristics that work for different groups’ interests, I put them together as a
domain labeled “groups’ interest”. Therefore. the final seven categories that I used in this
study were: economic and development issues. defense and foreign affairs. politics and
government organization, judiciary and order. care and communication (domestic).

women’s issues. and groups’ issues.

2. Coding and analysis

At the aggregate level of analysis. I make a general comparison of the level of
participation of the two gender groups within the various policy areas mentioned above.
In order to do this effectively. I coded the following items for each legislator: how many
initiations s/he proposed in each policy domain. how many times s/he chaired bills in
each policy domain, how many times s/he spoke in the committee meetings in each
policy domain, and how many times the s/he spoke in the Yuan meetings in each policy
domain. Because my primary interest is to differentiate the policy interests of each
legislator. all the bills proposed and discussed were coded to cover the entire range of
policy.

Furthermore, I adopted a relative measure at this stage so as to control the
differences between the groups and better apprehend how legislators distribute their time
and make the decision to give priority to a particular policy domain. Therefore, I code
the values of the variables (bill initiation. committee chairing. committee speech. and

Yuan speech) of each policy domain (A) as equal to the percentage of the total members

86

of either sex. In other words. I compare female and male legislators. as two distinct

groups, via the following sets of measurements:

(1) Initiation of issue A / Initiation of all issues (hypothesis 2a.)

(2) Chairship of issue A/ chairship of all issues (hypothesis 2b)

(3) Committee speech regarding issue A/ committee speech regarding all issues
(hypothesis 2c)

(4) Yuan speech regarding issue A/ Yuan speech regarding all issues (hypothesis 2d )

In comparing the policy priorities of the male and female groups by use of these
measurements. I expect to find that gender differences are most likely to be apparent in
two particular policy areas: namely, areas of traditional women's interest (education.
children. health. welfare. and communication) and feminist policy areas. Consequently. I
will pay particular attention to these policy domains.

In the individual level analysis. OLS regression is applied to each legislative stage.
The eight measures of my dependent variables are bill initiations. committee chairships.
committee speeches, and floor speeches. on the caring and communication issues and
feminist issues. In addition to the coding of individual legislators’ participation in these
policy areas as dependent variables, I also code independent variables. These include the
factors that influence the legislative participation of representatives. such as the
individual’s age. education. party identification, and level of constituency urbanization.
Since constituency is a major explanatory variable in legislative participation, it is
expected to have an affect on women’s issues. as urban representatives are more likely to
have experienced a diversity of lifestyles and pressure groups than are their rural
counterparts. Age is included due to Norris’s findings that younger legislators are more
likely to support women’s issues than older ones (Thomas and Welch 1991 ). Each
legislator's top level of education is also regarded as an important factor due to the
expectation that better-educated representatives are inclined to be more open to women’s

rights issues than less-educated ones. Party affiliation can also influence the level of

87

participation in issues. For example, DPP members are known for their enthusiasm for
political reform issues. Some might argue that opposition parties are more aware of, and
more open to, social welfare and social reform issues (such as women’s right issues.)

F inally. and taking the legislative process as a whole. I am also interested in
comparing gender difference in different legislative activities. In order to determine
whether gender differences vary in magnitude at different stages of the legislative
process, I compare the gender distributions evident within different legislative activities.

including the setting of agendas and policy deliberation.

Method 2: Interview study on legislators’ attitudes (i.e., hypotheses 3-5)

In order to test the attitudinal hypotheses 3 to 5. interviews were conducted. The
sample representatives consisted of all the female legislators (with the exception of one
who declined), and a near-equal number of male legislators (so as to be able to produce
an accurate comparison). serving in the Taiwanese National Legislative Yuan from 1995-
1998. In sampling the male legislators, I attempted to choose men who exhibited similar
party affiliations, ethnicity. levels of education, and district backgrounds to their female
counterparts. so as to isolate the affects of gender to as large a degree as possible.

Even with a matching sample. a legislator's perception of representation still
depends on various factors. Advance preparation regarding each legislator's background
was very useful as regards helping me interpret the significance of their responses and to
check for inconsistencies and. therefore. as mentioned above, I sought to become well
acquainted with each legislator's background. Following the setting-up of this data bank.
I sent letters to 52 legislators (a little more than the matching sample required. so as to
accommodate possible rejections). each requesting an interview. The letters (see

Appendix 2-a) described my research interests and the aims and intentions of this study in

88

a general manner.” Of the 52 legislators from whom I requested an interview. 40
accepted. essentially meeting my requirement for a matching sample. Most of the
interviews were conducted in legislators’ own offices or small conference rooms. and the
interviews took place between September 1997 and January 1998. during the fourth
session of the third term legislature.

At the beginning of each interview. I asked some general questions regarding how
deputies conducted their business. how they perceived their legislative role and function.
and which institutional position was. for them. important as regards gaining influence in
the policy-making arena. These questions elicited some rather general responses from the
representatives that, though they might move beyond my hypotheses. are important in as
far as they help to gain an understanding of how deputies legislate and represent, and thus
help dig the foundations of a more solid legislative study. Following this opening. and to
test hypotheses three to five (on gender difference). I elicited answers on two main
aspects of my study - the legislator's identification of their constituency, and their
individual policy priorities. I asked legislators to discuss which constituency groups and
policy issues were important for them to represent and the reasons why. Appendix 2-b
reproduces the interview questions. which are subdivided into several sections
corresponding to hypotheses 3 and 4, as outlined earlier.

Because half of the questions were designed to gain an understanding of legislators’
identifications and perceptions regarding legislative processes. and the issues involved
are thus complex. open-ended questions are often regarded as more useful than their
yes/no or highly structured counterparts. In addition. elite interviews such as these are
regarded as substantially different from the highly structured and standardized format of

survey research (Johnson and Joslyn 1991). For example. I am interested in an elite

 

’2 To avoid directing legislator’s responses. before and during the earlier part ofthe interview. I tried not to
Slgmfy the gender distinction. Without any anticipation. legislators are able to reply to the questions with
the” genuine reactions.

89

interviewee’s own interpretation of issues and events and. therefore, I do not want to
close off access to this potentially valuable source of information by constraining
responses. It is noted that “a less structured format is relatively exploratory and stresses
subject rather than researcher definitions of a problem”(Johnson and Joslyn 1991). In
practice. elite interviewees are sometimes reluctant to respond to a clearly standardized
set of questions.

Since most of the questions are open-ended. the categories within which I classify
the answers flow from the responses of the interviewee and refer to the theories only
when relevant. In short. I apply the interview data to test hypotheses three to five -
legislators’ perceptions of their constituency. their policy priorities, and the association
between these two. Finally, I also attempt to use the results of the interviews to generate
a possible explanation for the pattern of gender difference that I found in Taiwan’s

Legislative Yuan.

90

CHAPTER 4

LEGISLATORS IN TAIWAN:
A BEHAVIORAL ASSESSMENT

This aim of this chapter is to discover whether the descriptive representation of
female legislators guarantees the substantive representation of women’s interests. As
previously suggested. the descriptive representation of women refers to the presence of
women in political office. whereas the substantive representation of women demands
some form of actual activity by female deputies as regards the furthering of women’s
interests or concerns. In other words. I examine if the mere presence of female deputies
is related to the realization of female concerns, as the politics of group difference
assumes. I intend to test this hypothesis by discovering whether female legislators
behave differently from male legislators with respect to female group concerns, and I will
do this by examining two behavioral and three attitudinal hypotheses - the first two
behavioral ones in this chapter. and the attitudinal ones in the next chapter.

In the first hypothesis. I focus on the institutional placement of legislators in terms
of their selections regarding which committee to sit on. and the election of committee
Chairpersons across sessions. The second hypothesis will be organized into four stages:
bill initiation. chairing of committee meetings, speeches in committee meetings, and
speeches in Yuan meetings. In each of these sections. I will compare the actions of
female and male legislators in different committees and areas of policy. As described in

chapter 3. the time frame of both analyses is 1990 to 1997.

Hypothesis 1: Institutional placement of legislators: the committee system
Sheng (1996) argued that. in contrast to the situation in the United States. the

committee system did not play an important role in the policy-making process in Taiwan.

91

This was for a number of reasons. F irstly, bills can bypass committee deliberation and
move directly to the Yuan meeting if 20 legislators initiate this and a majority approve
the action. Even bills actually processed by the committees can go to the Yuan meeting
before they have been fully deliberated and, therefore. it can be seen that the authority of
committees can be easily superceded. Secondly, the final results of the committee
discussions are not treated as being more important than the original proposals or
amendments. Following the committee's discussion of a bill, their decisions are sent to
the Yuan meeting along with the original proposals, and all amendments are considered
equally. Moreover. no closure rules are applied in the committee. so committee reports
are not the final decisions. and legislators outside the committee can take part in the
committee's discussion and sometimes even make amendments. In fact. the only
differences between members and non-members of the committee are the committee
voting rights accorded to the former.

However, based on my interviews with legislators, most would appear to consider
committee discussion as very important if they want to influence and lead certain bills. It
is certainly the case that bills bypassing the committees are not very common - only when
the bills face time pressure or the level of revision is negligible will a full deliberation of
a bill by the committee be considered unnecessary. The legislators told me that, to a great
extent, committee assignments frame a legislator’s policy orientation and the groups he or
she will remain in contact with. And if the legislators consider some bills as important,
they will attend to the bills persistently through the whole committee process so as to
avoid unfavorable or unwanted revisions. While making appointments with legislators.
their assistants often told me that it was easier to schedule them on the day of a
committee meeting because legislators would come to the committees, whereas they
would often not appear for the floor meeting. Also, while waiting in the legislators’
offices prior to the interviews, their guests were usually those who dealt with issues

within the legislators’ committee jurisdiction. Admittedly, one reason for this is that the

administrative department liaisons are often in regular contact with, or even socialize
with. legislators’ assistants. particularly those assistants whose superiors serve on
committees related to their own subject area. Therefore, even though the committee
arrangement has not yet established a tradition of specialization and seniority in Taiwan’s
Legislative Yuan. it still provides a medium through which legislators’ can focus on the
concerns of their constituencies. In addition. the committee system helps to establish
their power bases and shape their policy focus. Therefore, given the significance of the
committee system, I sought to evaluate the impact of gender on the legislative structure
by examining both the legislators’ choices as regards committee membership and

Chairpersons.

Committee membership by selection

The issues that relate to a given legislator's choice of committee can be rather
complicated — affected as they are by various considerations. both individual and party-
related. The same legislator might join very different committees in different sessions for
different reasons. Therefore. it is not necessarily the case that a legislator's presence in
one particular committee will reveal their true policy interests to us. With the legislative
activity of each term spread across six sessions. only few of them choose to focus on a
single committee throughout (Sheng 1996). In other words. committee membership
choices are fluid rather than static.

In general, most of the legislators do not feel that they suffer any discrimination as
regards being able to join their choice of committee. A marginal hypothesis states that.
because of gender discrimination. women tend to become members of less prestigious
committees or those concerned with women’s issues. and men of more prestigious ones.
such as economic or local development. Yet this hypothesis does not seem to hold in

Taiwan’s legislatures. based upon the evidence contained in Chou’s previous study

93

(1990) and my own interviews. According to the legislators themselves. they can either
join the committee of their own choice or cast lots to decide who gets to choose.

Notwithstanding the lack of actual gender discrimination. I assume that gender does
shape the deputies’ personal committee choices. In my first hypothesis. I assumed that
female legislators were more interested in joining the committees dealing with women’s
interests - i.e.. education, welfare, children. health, and feminist issues. Most of the bills
dealing with such traditional women’s interests as the first four of these issues. are
referred to the interior and education committees; whereas the last of these - the feminist
issues - are dispersed across different committees, though they are usually referred to the
interior and judicial committee at some stage. It is thus assumed that women are more
likely to be members of the education, interior. or judicial committees, whereas men are
more likely to be members of economic and government organization committees.

I examine the gender division of labor as regards committee membership for three
terms in Table 4-1. Each representative can choose one committee in each session —
therefore, each member can make six committee choices across the six sessions. In Table
4-1, two forms of percentages are present in each cell, indicating: (l) the percentage of
women’s/men’s membership of each committee. and (2) the percentage of
women’s/men’s membership in all committees. These measurements show the relative
frequency of participating in any particular committee among two groups - women and
men.

In the first term, on average. female legislators as a group registered more often than
their male counterparts in four committees: education. interior. government. and
transportation. In the second term, female legislators registered more often than their
male colleagues in three committees: education, defense, and foreign. In the third term.
among all the committee choices, women registered more often than their male
colleagues in six committees: budget. defense, education, judicial, and transportation.

Therefore. across three terms. it is only in the education committee that female

94

Table 4-1. Committee memberships from 1990-1997 (3 legislative terms)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1990-1992 1993-1995 1996-1997
Female /Male/ Total F emale/ Male /Total Female/ Male /Total
Members 12 88 100 17 141 158 22 139 161
Budget ( count) 8 61 69 7 67 74 17 70 87
%within budget 11 89 100 9 91 100 20 80 100
% within gender 1 1 12 7 8 16 11
Defense 10 74 84 1 3 87 100 1 5 65 80
% within Def. 12 88 100 10 90 100 18 82 100
% within gender 14 15 13 1 1 14 10
Economics 6 56 62 7 101 108 l 1 71 82
% within Econ. 9 91 100 6 94 100 13 87 100
% within gender 8 12 7 12 10 1 1
Education 1 1 32 43 33 67 100 21 60 81
% 26 74 100 33 67 100 26 74 100
16 6 32 8 19 9
Finance 7 84 91 3 104 107 4 81 85
% 8 92 100 3 97 100 5 95 100
10 17 3 l3 4 12
Foreign 0 25 25 10 54 64 9 5 1 60
% 0 100 100 16 84 100 15 85 100
0 5 10 7 8 8
Government 7 40 47 5 70 75 3 67 70
% 15 85 100 7 93 100 4 96 100
10 8 5 9 3 10
Interior 15 53 68 9 99 108 5 76 81
% 22 78 100 8 92 100 6 94 100
21 1 1 9 12 4 11
Judicial 0 32 32 6 65 71 l 1 47 58
% 0 100 100 8 92 100 19 81 100
0 6 6 8 10 7
Transportation 7 42 47 8 96 104 1 3 75 88
% 14 86 100 8 92 100 15 85 100
10 8 8 12 12 11
Total 100 499 101 810 109 663
100 100 100 100 100 100

 

 

Note: Not included are the president. the vice-president, and those who served less than
one year in the Yuan. Two forms of percentages in each cell: the first row indicates the
percentage of women’s/men’s membership of each committee. and the second row
indicates the percentage of women’s/men’s membership in all committee

95

 

legislators consistently outnumbered male legislators. In contrast. in committees such as
those covering matters of economic and finance. male legislators always registered more
often than female legislators did. In other committees, no systematic pattern appeared
across the three terms. It should also be noted that. in those committees where women
outnumbered men. a positive relationship between the variable of gender and the
particular committee was also apparent. The opposite also held true. However, it is
important to stress that the resulting correlation was not hugely significant - only in the
second term were women and education significantly linked (0.302**).

I appear to have produced a slightly different set of results to Chou, Clark, and Clark
(1990). Their pre-1989 study found that women were more likely to serve on the
education and interior committees. Also, their OLS regression of gender on these two
committees (education and interior) proved significant after controlling for other factors.
though there were no statistically significant differences in any other areas. My 1990-
1997 study still shows that female membership of the education committee has
consistently exceeded that of men across the three terms - in fact, the education
committee has been the most popular committee for female legislators throughout this
time. Yet. the general pattern has lessened in its significance. In addition to women’s
consistent predominance in the education committee. my results also show several

changing patterns. Firstly. as regards the interior committees. gender and committee
membership were positively correlated only in the first term. By the second and third
terms. women had become less likely than men to join these committees. A possible
reason might be the growing importance of the interior committee in dealing with the
myriad political issues that followed democratization, however, in some traditionally
male—oriented committees. female visibility actually increased. An example of this would
be the fact that, in the second and third terms, female legislators were more likely than
their male counterparts to be members of the defense committee. Equally. in the

transportation committee. which is a popular one, there were on average more women

96

Co

.“7
LL".

.9,
.\\.

'9'
It.

.1.

ill:

registered than there were men in the first and second terms. This clearly demonstrates

that women have diversified.

Committee chairpersons by election

Taiwanese committee chairpersons carry similar institutional power to US
committee chairpersons in such areas as the calling of meetings, the arranging of
hearings, and the establishment of agendas. However, in contrast to the US Congress, no
seniority tradition exists in Taiwan’s Legislative Yuan. Indeed, studies of the Legislative
Yuan show that newcomers are elected as chairpersons more often than their senior
counterparts (Sheng 1996).

The number of female committee chairpersons represents a significant pool of
elected women officeholders with institutional leadership responsibilities. In the first
term (85-90 sessions). 16 out of 109 elected committee chairpersons were women. Of
these 16 incidences, women served as the chairpersons of the education committee most
often (6/16). and the interior committee next (4/16). In the second term, 21 out of 153
elected committee chairpersons were women, the largest numbers being elected to the
chairs of the education committee (4/21), and the foreign affairs committee (4/21).
However. the distribution of female chairships across the various committees has, in the

second term. become increasingly uniform. For example, three served as the defense.
three as the transportation. and still another three as the legal committee chairpersons. In
the third term, 18 out of 127 committee chairpersons were women, with the chair of the
education committee remaining the most common (5/18). and the judicial committee
lying a close second (3/18).

Across three terms. the gender patterns associated with committee chairship are
similar to those of committee membership. Women were most likely to join the
education committee and to be elected its chairperson, and less likely to join the

economic committee or be elected its chairperson. It is as well to note that. across the

97

.
v v-
“I".

U"

(is-

three terms, the number of women elected to the chair of the economic and finance
committee has not exceeded two. However, the gender difference has not been truly
significant across time. and the degree of difference in some other committees is not as
obvious as in this case. Overall. then, with very few exceptions gender appears not to
constitute a major dimension in the horizontal labor division, in terms of both committee
memberships and chairships. These are findings at odds with previous studies, and this
change has most likely occurred because of the new and expanded range of experiences
now open to political women.

The choice of a committee is a result of several factors. so much so that is difficult to
find. within it. a clear pattern of labor division by gender. F enno (F enno 1978) suggested
that there were three basic goals of all lawmakers that could affect the committee
preferences of legislators: namely. reelection. the gaining of influence within the house,
and effective public policy. He argued that these three driving motives would lead
legislators to seek certain different types of committee: the reelection-oriented (such as
Interior committee), the influence-oriented (Appropriation or Ways and Means), and the
policy-oriented (such as the Education and Labor committees). Since then, many other
scholars have expanded upon his views but all agree that some mix of these three goals
do indeed motivate matters such as committee choice.

The evidence above would tend to suggest that women would be divided by
different institutional and individual motivations, as well. Therefore, based upon my
observations and interviews with legislators, I suggest that there are two "levels" of
factors that dictate committee choice in the Taiwanese Legislative Yuan. Essentially,
these are: (1) the individual level. at which legislators’ committee choices are affected by
different interests. constituency bases. and personal benefits; and (2) the aggregate
(institutional) level. at which committee choices will also be subject to the affects of party
political maneuvering.

To expand on the first level. it is clear that committee choices are affected by

98

individuals’ interests derived from such diverse sources as formal professional training.
ingrained ideological concerns. personal profits. or campaign promises. Individual
deputies will rarely base their choice of committees to join entirely upon policy; rather.
their committee choices tend to accommodate multiple concerns. including both material
and symbolic interests they may have.

C learly. some legislators will choose to join a particular committee because
expertise gained within a previous profession provides them the leverage to participate
effectively. As a number of the legislators indicated to me. committee discussion
generally requires a more in-depth understanding of policy than does floor discussion.
For example. some representatives with teaching experience, being familiar with the
school system. might join the education committee; some of those with
business/professional backgrounds might join economic or financial committees; and
those with a civil-servant background might join the government organization committee.
Therefore. with most of the female deputies possessing educational backgrounds it is to
be expected that women would feel more comfortable about choosing the education
committee than any other. Consequently. the social division of labor can in a way be
seen to have transferred. institutionally. into the Legislative Yuan.

In addition. long-term ideological or personal concerns might affect an individual's
committee choices. For example, in the second term, a number of female legislators
joined the judicial committee because many feminist and children's bills were discussed
there. However. a feminist-oriented deputy will not necessarily manifest her ideological
commitment by joining a committee concerned with women’s interests. For example,
certain female legislators will purposely join committees that have no connection with
what are traditionally seen as women’s issues. The aim here would be to demonstrate the
capacity to work in a variety of fields. In fact. as one self-identified feminist legislator

put it.

99

Women are stereotypically oriented at the education and interior committees, and it
has been taken as the way it should be. Not only good men should serve the military
service, I think good women should too. In order to challenge the stereotype
imposed upon women. we should join the defense committee to let people see
women’s role in that committee. instead of being stereotypically restricted. And it
seems obvious that many female legislators have confined themselves to the
education committee. The educational committee has enough people already, so
they don’t need me. I would go to those committees where women’s visibility,
opinions. and perspectives need to be further explored (interview: 98022).

In addition. representatives might also base their choice of committees on individual
material benefits. After all. deputies come from different backgrounds and do themselves
have different needs. By joining certain "relevant" committees. legislators can push laws
along or pressurize the executive departments to make policy favorable to their interests
or their constituency’s interests. It is certainly the case that legislators will join
committees to “escort” their favorite bills through the Yuan - bills from the passing of
which they will personally benefit. In addition, some interest groups have close
associations with certain legislators, which can also directly or indirectly affect a
legislator's choice of committee. For example. legislators with business interests may
well seek to join the economic or budget committees so as to monitor their own interests.
Such materialistic motives are assumed to be equally rife within both gender groups and.
according to my interviews and some of the opinions expressed therein, this would
certainly appear to be the case. However. many female legislators also pointed out that.
unlike men. most female deputies are professional full-time legislators. i.e.. they seldom
have outside business networks or connections to protect. This could be another reason
why there are fewer women in the economic and budget committees.

Moreover. representatives tend to work for the interests of their constituency and to
fulfil their campaign promises. Therefore. when seeking re-election. members tend to
participate in the committees dealing with the bills upon which they campaigned as
joining such committees will enable representatives to claim to have lived up to their

promises. Legislators might also choose to join a committee in order to be seen to be

100

responding to the concerns of what they envisage to be their supporting constituency (re-
elected constituency.) F or example. as one female legislator put it: “some KMT female
legislators. such as A. B. C. and D, have been supported by the Huang. Fu-shin party
caucus (the KMT’s military electoral machine) to get elected. In earlier times. they
joined the defense committee to oversee the bills for their own constituency. However. as
time passes. they no longer want to stay at the same committee, instead they might try to
shop around [and experience] different committees.”(lnterview #98002.) This might
explain why the defense committee. which used to be a male-oriented committee in other
parliamentary countries. has been a popular committee for women in Taiwan.

Secondly. and at the aggregate level, the interests of the party can also guide
legislators’ committee choices. This often happens in two ways: by the party urging its
member to move to a committee dealing with the party’s concerns. or by persuading its
members to attend less popular and prestigious committees. On the one hand, parties are
always looking to promote certain issues or bills. and they will therefore try to ensure that
these bills are carried through by coordinating the actions of their members as far as
possible. For example. if in one legislative session a particular party has several
important welfare bills on its agenda. the party will attempt to guide its members onto a
related committee (e. g.. the interior committee), so as to better influence the outcome of
the legislative process. It is common for parties to direct most of their members to one
particular committee if they have important agendas to promote in that committee's
subject area during any given session. This "clustering" of one party’s members in one or
two committees is most common amongst the small parties. because they will often need
to gather together in order to obtain sufficient critical mass to attempt to further their
agenda. This, indeed. can obtain for the minority parties a certain degree of leverage in

deciding committee agendas and bill-proceedings." and yet. sometimes. it may well

 

"’ For example. in the second term ofthe legislators. the NP mobilized all oftheir members tojoin the
National Defense committee in the fourth session. and the Judicial committee in the fifth section. hence it

101

obstruct an individual member’s personal interests. One New Party legislator (interview:
98024) describes the constraints that surround the choosing of committees to join as “the
pity of the New Party.” He says that, given the choice, he would attempt to join A or B.
but that, for the past three sessions, he has disregarded his own choice of committee in
favor of joining that which best serves the interests of his party. “Because for a small
party. if the party wants to lead some agendas today. it must mobilize its members to the
one committee. And once it chooses to do so. its members lack the autonomy in
choosing the committee”. Similar comments were heard from another NP female
member (interview: 98017), who had been assigned to some totally unfamiliar
committees. instead of her preferred and familiar committees.

On the other hand. the party whips might channel members to some less sought after
committees. such as the foreign affairs or judicial committees, ifcompetition for some
popular committees, left unchecked. would entail them exceeding their quorum (the
maximum size is 18 members per committee). For example, the party may redirect some
members to their second or third choice committees in certain sessions, and in return
promise them their first choice in the next session. Contrary to examples of the clustering
effect, this need to urge party members into less popular committees is most likely to
occur in the majority party. and members will invariably cooperate in return for some
form of future gain. In the course of my interviews it became clear that female deputies
are as subject to these pressures as male ones: forced to attempt to balance the interests of
their party with their own interests. but often left with little choice but to bow to the
influence of the party. For example. one KMT female legislator (interview: 98038) told

me that she once intended to participate in the interior committee. however, her party

 

has two chairs and one schedule member in the National Defense committee; and one chairperson and one
schedule member in the Judicial committee in the fifth session (Sheng 1996). DPP are more concerned
about the political bills since the party was formed. they have more of their members in the political affair
committees (interior and government organization).

102

asked her to join the less-popular foreign affairs committee. In summary then, it is clear
that the committee choices of female deputies are. to a certain degree, influenced by the
wishes of their party.

Nevertheless, generally speaking, most of the legislators consider themselves to have
an essentially free hand in choosing the committees they would like to join, at least in
some sessions. However. in addition to the influences of party, their choices are also
shaped by their expertise. experiences. bargaining power, interests, self-interest. and re-
election considerations. Clearly then, there are a large number of different elements
involved in choosing a committee to join and, consequently, a clear gender division as
regards committee arrangements is extremely difficult to discern. Although female
legislators would appear. generally, to prefer committees like education. the effects of

such preferences cannot be viewed in isolation.

Hypothesis 2: Participation in legislative activities

Stipulated committee memberships cannot tell us much about the policy subjects in
which individual legislators are really interested or the degree to which they engage in
policy representation. Because committees do not have clear jurisdictional
compartments, most policy subjects overlap committee jurisdictions, enabling members
to master several policy fields and enhance the interaction and coordination of each
committee in turn (Davidson and Oleszek 1996) but often obscuring a member's principal
policy interests. For example. the interior committee deals with various matters,
including political issues. civil affairs. elections. labor, welfare. health. and the police.
Members of the same committee might therefore focus on very different policy subjects.
and committee titles certainly do not give a clear indication of their substantive purview.

In addition, the question of which committee a legislator joins is not sufficient. in
itself, to discern the actual and effective level of participation by the representative in the

policy-making process. As Hall (1996) put it:

103

By focusing on the formal institutional units, we too-conveniently label as
institutional or structural something that is better understood as a behavioral
phenomena [albeit subject to institutional constraints]. Taken literally, at least. the
view that committees constitute a system by which the organization divides labor -
thereby achieving specialization. expertise, and greater organizational efficiency -
founders on an important behavioral puzzle. Legislative labor in Congress is
required of no one. Few specific tasks are assigned; no production quotas are set.
No sanctions await senators and representatives who show up late, disappear when a
constituent call or the golfing weather is good, or otherwise labor little in legislative
decision making (Hall 1996, 11).

As Hall noted. the institutional arrangements serve as a point of departure but not as
a satisfactory indicator of the level of policy responsiveness and participation of the
legislators. As he points out. serious actors tend to be members of committees with
jurisdiction. yet, members of such committees need not to be serious actors. Therefore. it
is necessary to delineate the institutional and behavioral divisions of labor in the policy-
making process; in short. we must move beyond the merely institutional arrangements.
Therefore. I now turn to the dynamic processes of legislative participation. as stated in
hypothesis two.

To evaluate (as suggested in the previous chapter) policy representation as a whole, I
break down the legislative process into two phases: the agenda setting phase and the
policy deliberation phase. F urtherrnore, agenda setting in the Legislative Yuan generally
involves two stages: first. deciding which issues get to be addressed at all, and this is
determined by the bill-initiation; second, deciding which issues get to be examined first
in the committees, and this is determined by the chairperson's arrangements. In the
beginning. legislators usually define their legislative agendas and the issues to which they
will give effort and time. Next. policy deliberation is a long-term process proceeding
both in the committee and Yuan meetings. Once the bills are on the agenda, it is then the
speeches made during the policy debate and the deliberations that follow that determine
the final result of the bills. Therefore, at each legislative stage (i.e.. for each of these

measures) - initiation. chairship. committee speech. and Yuan speech. I will examine the

104

policy interests and representation by gender.

Stage I: bill initiation ( hypothesis 2.a)

Various policy issues occupy an individual legislator's time and attention. and the
issues that are addressed at the first stage of the process set the scene for what will follow.
Hypothesis 2a assumes that female representatives are more likely than their male
counterparts to initiate bills related to areas of women's issues and. to test whether such a
gender pattern does indeed exist. I examine all the bill-initiations in the Legislature Yuan
through the three terms of my observation.

By designation. bill-initiation in the Yuan meetings requires the co—sponsorship of
15 legislators besides the original initiators. Generally speaking, most of the bill
proposals or amendments are drafted by one legislator’s office even though the bills
might have multiple authors. As I explained in the previous chapter (footnote no. 4). the
act of co-sponsorship is not considered a particularly meaningful measurement of
legislative participation and. therefore, all the bill initiations analyzed here are based
solely on the first authors. For bills with multiple sponsors. the initiator is determined by
reference to the documentary records, namely, a monthly report on bill-initiation.
provided by the secretary of the Legislative Yuan. I examine the gender patterns by
dividing the whole issue of bill initiation into seven policy subjects as referred to in the
methodology section. Table 4-2 shows this aggregate data ofbill initiation arranged by
genden

As shown, the discrepancy between the first and second terms as regards the total
number of bill initiations is caused by the large number of tenured members that sat
throughout the first term, performing no law-making function and all retired by the
beginning of the second term. Consequently. the dramatic increase in the total number of
bill initiations indicates that legislators have been increasingly engaged in active

legislation since this occurred. On average. in the first term. female legislators initiated

105

Table 4-2

Numbers and proportions of bill sponsorship of each policy area

 

1’" term 2“d term 3rd term
Female Male Female Male Female Male

 

Economic
Development 2(7%) 320.2%) 27(250/0) 151(28%) 12(17%) 173(360/0)
Defense and

 

 

 

 

Foreign 1(3%) 4(3%) 3(3%) 10(2%) 1 (1 %1) 22(5%)
Politics and
Govern 3(10%) 53(37%) 16(15%) 168(32%)1(1%) 93(19%)
Judicial and
Order 4(13%) 17(12%) 16(15%) 48(9%) 14(20%) 61(13%)
Caring and

Communication l4(47%) l8(l3%) 21(20%) 80(15%) 29(42%) 81(l7%)
Feminist

 

5('17%) 2(1%) 16(1 5%) 10(2%) 9(1 3%) 3(0.7%)

 

Group
1(3%) 17(12%) 8(7%) 63(12%) 3(4%) 51(10%)

 

Total

30 143 107 531 69 486
Cell entries are the numbers of bills initiated in each policy area. Figures in parentheses
indicate the ratio of that type of bills among all policy area within each gender group

 

Source: The monthly legislation reports (primary source) and the Legislative Gazette.

106

2.5 bills (30/12) and male legislators initiated 1.7 bills (143/86). In the second term,
female legislators initiated an average of 6.3 bills (107/17) and male legislators. 3.8 bills
(531/138). In the third term. the figures were 3.1 (69/22) and 3.5 (486/137)” for females
and males respectively. Given the very limited number of women in the Yuan, the
differences here cannot be overstated without first examining the variation on a more
individual level. However. at first glance. it does seem that the conclusion of Chou,
Clark. and Clark (1990) that women deputies were less active in initiating bills in
Taiwan’s legislature is no longer valid. Even at this rather early stage it already looks as
if women are now playing an active role in the making of laws.

The aggregate data allows me to demonstrate the different policy preferences
between women and men. For men. in the first and second terms, the first priority of bill
initiation lay in the public field - dealing with political issues and government
organization; and the second priority lay in the field of economic development - dealing
with issues related to economics, finance. transportation, and local development.
However. by the third term. this order of importance appears to have been reversed. For
women, in the first and third term, the top priority of their bill-initiations concerned
matters of caring and communication. Whereas, in the second term, women initiated bills
more evenly across different policy areas. with the number of bill-initiations in the area of
economic and development policy slightly exceeding that in the caring and
communication area. With an absolute majority in the Legislature Yuan, the men have
consistently tended to emphasize political and economic issues, whereas women have
tended to focus upon the issues of traditional interest to themselves.

In the policy domains of economic development and politics and government, male
legislators consistently outnumbered their female counterparts in terms of bill initiation

across the three terms. In the first and second term. male legislators initiated

 

1" Bill-initiation here includes the sponsorship on the draft ofa new bill. the major amendments. and small
amendments.

107

approximately one-third of bills concerned with these areas, in which matters of political
power and economic resources lie. Perhaps because of the public/private dichotomy. and
thus their traditional exclusion from these issues. women had less experience and interest
in investing their time in these areas.

When examining these differences in policy areas of female interest. the gender gap
becomes even more apparent. In the first term. of the 30 bills initiated by women, 47
percent were related to the caring and communications area and 17 percent to feminist
issues. However. of the 143 bills initiated by men. the figures for these areas of policy
were only 13 percent and 1 percent respectively. In the second term. and as the level of
female bill—initiation increased dramatically, of the 107 bills proposed by women, 20
percent were proposed in the area of the caring and communication, and. this time, 15
percent concerned feminist issues. On the other hand, of the 531 bills proposed by men
in that term, the figures were 15 percent and 2 percent. F inally. in the third term. of the
69 bills proposed by women, 42 percent were proposed in the domain of caring and
communication, and 13 percent in the feminist domain. The figures for men were very
different, with 486 bills proposed, but only 17 percent concerned with caring and
communication, and only 0.7 percent concerning feminist issues. Except in the second
term, it can thus be seen that the gap between the level of female and male bill initiations
in the caring and communication policy domain is substantial.

In the second term, though the subject of caring and communication is no longer
women’s first priority when it comes to the initiating of bills. they still placed more
emphasis upon this area than did men. In the first and third terms. than one-third of all
female bill-initiations concerned this policy area. Consequently, it can be seen that,
across three entire terms, women deputies consistently gave more weight to the subject
areas of caring and communication and feminist issues than did their male colleagues. at
least in terms of the relative level of bill-initiations.

Moving to the individual level by examining correlation and regression the

108

distinction becomes even clearer. The unit of analysis in the following six models is the
individual legislator, and I use their bill-initiations in the areas of caring and
communication and feminist issues respectively as my dependent variables across the
three terms. To examine exactly what it is that affects legislators’ policy preferences, I
design each model to include several independent measures so as to capture the extent to
which legislators’ policy priority is affected by gender after controlling for other relevant
factors. The independent variables include gender. age. education. the level of
urbanization of their constituency, and their party affiliation, as described in the method
section. The variables for gender and party are defined as dummy variables, with female,
party-DPP. party-NP. and Independents all set at 1. The level of district urbanization is
measured by the percentage of the total employed population in the legislator’s district
that works in the field of agriculture. For those party-list and other non-district
members, I substitute this statistic with the corresponding national figures. All these
independent variables were labeled in the methodology section as the general factors
influencing the level and type of legislative participation of the individual member.
Table 4-3 shows the results from OLS regressions.

It can thus be seen that gender is positively correlated with the level of bill
initiations by legislators in the caring and communication and feminist areas. In the
caring and communication policy area. and in the first and third terms, females are
significantly more likely to initiate bills. Though it should be noted that this is not the
case in the second term - in which age. education, and party were more significant in
explaining differences in the level of bill-initiation - it is consistent with the results of the
aggregate level analysis. Clearly, the policy interests of female deputies have become
more diversified, and the gender patterns are thus less clear in the area of caring and
communication. However. in feminist issue. gender is indeed shown to be a significant

factor in explaining the differences between male and female bill-initiation across the

109

OLS models of policy initiation

Table 4-3.

 

1" term

Care/Comm Feminist

2"d term

Care/Comm Feminist

3“l term

Care/Comm Feminist

 

Independent

variables
Constant

Age

Education

Gender

Rural

Party-DPP

Party-NP
Indep.

Model

diagnostics

Adjusted
R-square
F
N

4083
(407)
4068
(084)
.165*
(075)
.980***
(211)
4726
(666)
.044
(177)

.213
(.265)

.207

5.213***

98

.022
(168)
.0013
(035)
.0026
(031)
.386***
(087)
.026
(275)
4058
(073)

-.039

(275)

3.751**
98

.534
(488)
4273**
(105)
.282**
(089)
.410
(276)
42461**
(892)
.547**
(186)
L171**
(390)
4106
(535)

7.375***
155

.475*
(235)
4093
(051)
4032
(043)
.805***
(133)
4204
(430)
.101
(090)
.133
(188)
4057
(258)

7.283***
155

4420
(701)
.131
(158)
.128
(123)
.708*
(352)
-1135
((437)
.339
(263)
.845*
(396)
4592
((086)

.048

2138*
159

4059
(135)
4016
(030)
.03
(024)
.366***
(068)
4210
(277)
.042
(051)
.247**
(076)
4014
(210)

.229

7.690***
159

 

The President and Vice-President of the Yuan are not included in the sample N
***p<=.001

*p<=.05

**p<=.01

Source: 1. The monthly legislation reports. 2. The Legislative Gazette.

3. The Electoral Bulletin ofCentral Election Commission. 1989. 1992. 1995.

110

entire three terms. In essence. it would indeed appear that women tend to be more

engaged in the initiation of women’s issues than do men.

Stage II: committee chairship (hypothesis 2.b)

Once the bills are proposed and the first reading (report) takes place in the Yuan
meeting, the next question to arise is whether the bill will be discussed in the committees
or not. With hundreds ofbills initiated by legislators and the administration. not all bills
can possibly gain an equal degree of attention. with some bills remaining “frozen” for a
considerable period of time because insufficient effort has been put into getting them onto
a committee's schedule. Most legislators agree that the chairperson of each committee
possesses greater power than others in the decision-making process due to their
jurisdiction over the setting of their committee's agenda. One legislator stated that the
power of the chairpersons as regards agenda setting is absolute. and that this was
especially so for DPP and NP chairpersons, who can always place their own bills at the
top of the list (interview: 98008). Yet, for some KMT chairpersons, who also need to
consider the demands and preferences of the executive branches, the agenda setting
power may not be as absolute as it is for minority party chairpersons. Also, as noted, the
committee chairperson decides which bills get on the agenda for examination in the
committee meetings and, therefore, because of the critical position of the chairperson in
the legislative process. parties mobilize in order to get their own members elected as
chairpersons and leading the discussions.

Hypothesis 2b assumes that female representatives are more likely than their male
counterparts to chair committee meetings examining bills related to women’s issues. I
discussed committee membership and chairship in a more static sense, earlier, as regards
stipulated categories and without distinguishing between the respective policy subjects on
which legislators within each committee focus. Here, I turn to a dynamic approach to

investigate the gender pattern in terms of the policy interests of committee chairpersons

lll

Table 4-4:

Numbers and proportions of bill chairship of each policy area.

 

1" term 2"d term 3'" term
Female Male Female Male Female Male

 

Economic
Development II(I30/0) 95(25%) 22(23%) 184(34%)12(19%) 110(36%)

 

Defense and

 

 

 

Foreign 0(%) 8(2%) 5(5%) 13(2%) 2(3%) 14(4%)
Politics and
Govern 14(l6%) 73(20%) 8(8%) 121(22%) 5(8%) 65(2l%)
Judicial and
Order ll(13%) 64(17%) ll(l2%) 61(11%) 11(l7%) 54(18%)
Caring and

Communication 22(25%) 67(l8%) 41(43%) ll7(21%) 24(38%) 36(12%)

 

Feminist
2( 2%) 0(%) 6( 6%) 8(2%) 6(10%) 3(1%)

 

Group
28(320/6) 45(120/6) 2(2%) 44(8%) 3(5%) 25(8%)

 

Total
88 374 95 548 63 307

 

Cell entries are the numbers of committees chaired in each policy area. Figures in
parentheses indicate the ratio of that type of bills among all policy area within each
gender group.

Source: The index system of the legislative participation (primary source) and the
Legislative Gazette

in actually setting the agendas. Every time the committee meets, the chairperson will
usually schedule one bill for examination." Which bill is put forward. and by whom. is
my primary interest here. and. by dividing the bills discussed in the committees into
seven main categories (the same scheme used earlier and explained in the methodology
section). the gender patterns regarding the policy agendas of the chairpersons are shown
in table 4-4.

In the first term, 462 committee meetings on bill-examination. hosted by 8
chairwomen and 47 chairmen. took place. On average, chairwomen held 11 (88/8)
committee meetings and chairmen held 7.6 (374/49). In the second term, there were 639
such committee meetings hosted by 13 chairwomen and 82 chairmen. On average.
women chaired 7.3 (95/13) meetings and men chaired 6.7 (548/82). In the third term, 370
committee meetings were hosted by 10 chairwomen and 64 chairmen and. on average.
women chaired 6.3 (63/10) and men chaired 4.8 (307/64) meetings.

In the first term. 25 percent of the 88 committee meetings hosted by female
chairpersons were related to caring and communication bills. whereas male legislators
hosted a total of 374 committee meetings. with 18 percent of those they hosted examining
such bills. In the second term, 43 percent of the committee meetings women hosted
examined bills that concerned the above arena, compared to 21 percent of the committee
meetings that men hosted. Finally, in the third term. 38 percent of the committee
meetings hosted by women discussed bills that concerned caring and communication.
whereas only 12 percent of male-hosted committee meetings did the same. Thus. across
three terms. and as a group, female chairpersons were more likely than male chairpersons
to put bills concerning traditional female interests and feminist issues, on committee

agendas, and thereby increase the chances of the bill being noticed and passed. Also.

 

"’ Committee meetings have different functions. including : public hearing. interpellation, or bill
examination. Bill examining before the second reading is the most important function of the committee. In
each committee meeting, usually. the chairperson would schedule one bill for examination. In very rare
cases chairperson might choose two bills in one meeting.

113

women chairpersons appeared most likely to arrange for bills concerning caring and
communication issues to appear on committee agendas.

Contrary to this. male chairpersons were most likely to choose bills of economic.
finance. budget. and transportation issues (the domain of economic and development
issues) to place on committee agendas. In the first term, 25 percent of the committee
meetings hosted by men examined policies of that type. whereas only 13 percent of the
meetings hosted by women did the same. In the second term, 34 percent ofthe
committee agendas set by chairmen were in that policy area as opposed to 23 percent of
the committee agendas set by chairwomen. And in the third term. 36 percent of the
committee meetings chaired by men discussed such bills. The figure for chairwomen was

19 percent. Therefore, across three terms, it was generally the case that almost one-third
of the committee meetings chaired by male legislators discussed economic and
deveIOpment issues. The second most popular policy area for male-chaired meetings to
discuss was that of political issues and government organization. Therefore, once again,
men consistently emphasized "male issues" with respect to the scheduling of bills in the
Committee meetings.

However. even with these clear differences of preference. in terms of arranging bills.
betWeen chairmen and chairwomen, the factors that can affect a chairperson's choice of
agenda can still be varied. Even though a chairperson's preferred bills tend to be at the
top of their committee agenda list. some legislators (especially members of the majority
party) commented that they would also take pressure from the party and the
administration into account when arranging such agendas. Therefore. to examine exactly
Which factors affect an individual’s agenda setting in the committees. a further correlation
and regression analysis must be conducted. In this case. the unit of analysis is the
individual chairperson. I regress such individual characteristics as the gender, age,
CdLlCation, the level of urbanization of their constituency. and their party (three dummy

variables Of DPP. NP. and INP. as before) of the chairpersons on the committees they

114

 

 

chair for caring and communication bills and feminist issues respectively. With two sets
0 f dependent variables - the frequency of arranging bills concerning caring and
communication, and concerning feminist issues - for each term. six models across three

legislative terms are produced in Table 4-5.

First. the linear association between women and chairship in caring and
cornmunication areas is positive. and. except in the first term. significant. The difference
in this first term might have been caused by the small number of chairpersons - 8/12 - and
the vast number ofcaring and communication-related bills that were presented.
Alternatively. it might be due to the fact that there was still a system of group
representatives for educational matters at that time. and that male deputies were thus still
taking the lead in educational issues") The correlation between gender and chairship as
regards feminist issues is also extremely positive -— the regression analysis showing that
gender is a significant factor in explaining the feminist bill-setting agenda. after

controlling for other factors for the first and third term.

\

16
Before 1992. beside the representatives from the district election, there was still the system of group
lepreSentatives. under which certain candidates ran for the occupational list. instead ofdistrict list. They
22:: different constituencies. depending on their occupational constituency---business. industry. labor.
distri u ItUre. fishermen. and educatlon. Voters must choose whether to reg1ster to vote for wh1ch. llst-w
conceCt 0r occupat1onal. Occupatlonal group representatives tend to lead 1n the issue area ofthe1r group
1‘11 - lfI remove the two members ofthe educatlonal group representauves from the sample. gender

690 . . . . . . . . . . . .
1550 m e3 3 Sigmficant factor In explammg the d1fference 1n chamng the committees related to domestlc
es - "

115

Table 4-5.

OLS models of committee chairship

 

 

 

1” term 2"d term 3rd term
Care/Comm Feminist Care/Comm Feminist Care/Comm Feminist
Independent
variables
Constant -1.650 -.0.053 1.598 .738 .893 .649
(2.465) (.206) (1.563) (.393) (1.392) (.512)
Age .212 -.0092 -.350 -.153 -.289 —.185
(.578) (.048) (.363) (.091) (.297) (.109)
Education .841 .023 .381 -.045 .158 .044
(.442) (.037) (.291) (.073) (.231) (.085)
Gender 1.150 .254** 1.611* .178 1.677* .475*
(1.259) (.105) (.828) (.208) (.582) (.214)
Rural -3.081 -.015 -3.986 -.635 -.571 -1.229
(4.329) (.362) (2.865) (.721) (2.609) (.960)
Party-DPP -1.131 -.0005 -.156 .161 -.042 -.089
(1.279) (.107) (.575) (.145) (.430) (.158)
Party-NP -.560 .622 .737 -.140
(1.395) (.351) (.662) (.244)
Indep. -1.421 .0016 5.155 -.107 -.515 -.105
(1.696) (.142) (2.714) (.683) (1.757) (.647)
Model
diagnostics
Adjusted .129 .1 18 .063 .036 .099 .064
R-square
F 1.235 1.110 1.896 1.508 2.146(*.051) 1.713
N 57 57 95 95 74 74
*p<=.05 **p<=.01 ***p<=.001

Source: 1. The index system of the legislative participation. 2. The Legislative Gazette.
3. The Electoral Bulletin of Central Election Commission. 1989, 1992. 1995

116

Committee chairships can be seen as the link between legislators’ sponsorships and
speeches. Next. I examine whether the gender patterns of the speeches legislators made

are consistent with the above.

Stage III: speech in the committees (hypothesis 2.c)

Once the agenda in the committee meeting is set. the stage is set for the speeches.
Making a speech requires an individual’s time, effort. interest, and expertise. especially at
the committee stage. As discussed earlier. even though the committee examination is not
the most definitive or important stage of the law making process. it is still regarded as
vital. In committee meetings, it is usually the case that only a few members will attend
and, therefore, only a few people could end up deciding upon the main contents of the bill
that will be sent on for floor discussion. Non-committee members can also participate in
this process. however. they do not possess a vote if any changes are made to the bills.
Considering the fact that many bills, after examination at the committee stage, do not
experience a significant degree of alteration in the Yuan floor meetings. it can be seen
that the committee discussion can decide the entire direction and content of many bills’7
and is thus of huge importance.

Given the scheduling conflicts that often arise. legislators need to determine in
which committee meetings to invest their time. Hypothesis 2.c assumes that female
representatives are more likely than their male counterparts to speak in committee
meetings in which bills relating to women’s issues are discussed. Therefore. in order to
unearth any distinctive gender patterns in this process, I here classify the committee
speeches by policy area (using the same scheme as above) and whether male or female

representatives delivered them in Table 4-6.

 

'7 Especially in the third term. given the establishment ofthe mechanism ofthe party coordination. most of
the bills were not discussed on the floor. but rather in the committee and the informal meetings (party
coordination)

117

Table 4-6

Numbers and proportions of committee speeches made in each policy area

 

1" term 2"" term 3rd term
Female Male Female Male Female Male

 

Economic
Development 605(19%) 9882(30%199(9%) 3253(29% 225(22%) 2168(32%)

 

Defense and

 

 

 

Foreign 110(3%) 1889(6%) 75(3%) 213(2%) 45(4%) 313(5%)
Politics and
Govern 659(200/6) 9564(29% 242(11%) 2740(24%110(11%) 1533(230/0)
Judicial and
Order 361(11%) 4299(130/6 305(13%) 1149(10% 225(22%) 1216(18%)
Caring and

Communication 1196(37%) 4678(14%1284(56°/o) 2865(26) 244(24%) 902(13%)

 

Feminist
12(0.3%) 7(.0002%)96(4%) 159(.5%) 87(8%‘) 58(.9%)

 

Group
313(10%) 3098(9%) 75(3%) 843(8%) 89(9%) 612(90/0)

 

Total
3256 33417 2276 11222 1025 6802

 

Cell entries are the numbers of speeches made in committee in each policy area. Figures
in the parentheses indicate the ratio of that type of bills among all policy area within each
gender group

Source: The index system of the legislative participation (primary source) and the
Legislative Gazette

118

In the first term. on average. women spoke 271.3 (3,256/12) times and men spoke
388.6 (3.3417/86) times when examining bills in committee meetings. In the second
term. women spoke 133.9 (2.276/17) times and men spoke 81.3 (1 1.222/138) times in the
same circumstances. And in the third term, women made an average of45.6 (1025/22)
speeches and men made 49.6 (6.802/137) speeches in the committee meetings.
Compared to the number of speeches made in the Yuan floor meetings (to which I will
return later). on average. legislators tend to be more active at the committee stage.

I first compare the different subject emphases of these committee speeches between
women and men and across the three terms. According to this analysis. it was
consistently the case that women most frequently chose the policy area of caring and
communication upon which to speak; and that men most frequently spoke of matters
related to economic development. (Followed, except in the second term, by speeches
about political and government policy.) It would appear, therefore, that women tended to
address women’s issues, whereas men chose to address men’s issues.

Across the three terms. women consistently placed more weight than men upon
caring and communication matters and feminist issues. In the first term. of the 3.256
speeches women made in bill-examining committee meetings. 37 percent addressed bills
related to the caring and communications area. For men. of the 33,417 speeches they
made. 30 percent addressed matters of economic development area and 29 percent dealt
with political and government affairs. In the second term. the differences grew even
more apparent. with 56 percent of female committee speeches addressing bills concerning
caring and communication. but only 26 percent of male speeches dealing with the same
area. In the third term, though the gender differences remained, the size of the gap
decreased somewhat. Women still spoke most frequently in the area of caring and
communication, using 24 percent of their total speeches in order to do so. but the number
of speeches concerning other policy areas also increased. Turning to the area of feminist

interests. the committee speeches made by both men and women are few and far between.

119

especially in the first term. For example. in the first term. only one feminist bill was
scheduled for examination by a chairwoman.

Once again. I regress the individual’s party. age. level of education. degree of
constituency urbanization, and gender on committee speeches that address caring and
communications issues and feminist issues. The unit of analysis is. as ever, the
individual legislator. The regression results of six models are shown in Table 4-7.

Clearly. the correlation between gender and the number of committee speeches made
concerning bills related to caring and communication matters is positive across three
terms. but not always significant. Nevertheless, gender appears to be an important
variable in explaining such differences. Across three terms (i.e.. with the exception of the
first term). women were significantly more likely to speak upon bills that concerned
caring and communication issues than men were. Turning to the area of feminist issues,

the relationship between gender and that subject area is even stronger.

I“ /' >41 ——. ._,_.

S1

OLS models of committee speeches

Table 4-7.

 

 

1" term 2"d term 3rd term
Care/Com Feminist Care/Com Feminist Care/Com Feminist
Independent M W
variables
Constant -11.271 .485 8.853 2.209 -3.691 -.906
(60.358) (.418) (20.586) (2.331) (4.640) (1.183)
Age -12.655 -.097 -6.590 -.230 .130 .019
(12.511) (.087) (4.424) (.501) (1.046) (.267)
Education 33.528** .0097 8780* -.041 2.134** .215
(11.076) (.077) (3.758) (.425) (.816) (.208)
Gender 49.143 .853*** 49.072*** 3.593** 4.472(.057) 3.485***
(31.362) (.217) (11.632) (1.317) (2.332) (.595)
Rural 442.554 -1.124 -28.948 -6.423 -9.635 -1.914
(98.845) (.685) (37.643) (4.262) (9.507) (2.424)
Party-DPP 55.351 * .066 11.094 1.628 5.846*** 1.236**
(26.291) (.182) (7.862) (.890) (1.743) (.445)
Party-NP 47.279** 6.492*** 6.606* 1.782**
(16.450) (1.862) (2.620) (.668)
Indep. -37.039 -.06 -.258 -.297 -2.173 -.091
(39.292) (.685) (22.557) (2.554) (7.189) (1.833)
Model
diagnostics
Adjusted .148 .157 .191 .131 .138 .230
R-square
F 3.807** 4.005*** 6.123*** 4.263*** 4.621*** 7.738***
N 98 98 155 155 159 159

 

Note that the president and vice president of the Yuan are not included in the sample N
*4"? —-
p<—.001

*p<=.05

**p<=.01

Source: 1. The index system of the legislative participation.

2. The Legislative Gazette.
3. The Electoral Bulletin of Central Election Commission. 1989. 1992, 1995

Stage IV: speech in the Yuan meeting (hypothesis 2.d)

Following examination ofa bill by committee, the second and third readings of the
bills are conducted at a Yuan floor meeting. Most bills. before being sent to the Yuan
floor meetings. have. of course. been intensively discussed in the committee meetings.
Therefore. if no severe controversy or political sensibility is involved. some bills in the
second and third readings may not evoke intensive discussion.

Each committee elects two procedural committee members. who report and arrange
the order of the bills to be put on the Yuan agenda on behalf of their committees and their
parties. Sometimes personal preferences play a part. too. Twenty procedural members
work collectively to decide which bills will top the agenda. In the third term, and in order
to save time and increase the efficiency of bill proceedings in the Yuan meetings, a
mechanism of party coordination was established. This was necessary because both
because the workload of the lawmakers has increased dramatically in recent years, and
because competition among multiple parties (three parties and some small factions) has
made the Yuan meetings inefficient. Legislators are, therefore, encouraged to participate
in informal party coordination" to solve disagreements and attempt to strike a common
ground before bills are actually sent to Yuan meetings. Such informal meetings
determine both which bills will get addressed and which version of the bill will get
precedence in the Yuan meeting. Consequently. discussions in the Yuan meetings have
become less fierce and energetic. since most of the disagreements have already been
settled.

However. the chances ofa bill being passed are still, to a large extent, determined by
the second and third readings. Legislators with different viewpoints need to make their
presence known and express their concerns. Also, for non-committee members. Yuan

(floor) meetings provide the chance to make meaningful revisions to a bill.

 

'8 The original informal party bargain and coordination is now under consideration to formalize it.

Legislators speak upon various issues with differing levels of intensity. Generally
speaking. the more a legislator addresses one particular policy area. the more he or she is
obviously concerned about that issue. even though we cannot discern their position or
approach from such a simplistic measurement. Hypothesis 2d assumes that female
representatives are more likely than their male counterparts to speak during floor
deliberation upon issues related to women’s interests. I must therefore examine
differences in the frequency of speeches made by women and men among the seven
policy subjects in the second and third readings across the last three legislative terms as
shown in Table 4-8.

As a whole. in the first term. women spoke an average of 40.7 times (488/ 12) and
men spoke an average of 49.7 times (4.277/86) at the floor meetings of the Legislative
Yuan. In the second temi. women spoke an average of 32.5 times (553/ l 7) and men an
average of23.4 times (3234/138). And in the third term. women spoke 23.4 times
(23 3x22) and men spoke 9.4 times (13207137).

The different policy preferences as regards speeches made in floor debates between
men and women appear significant. Female legislators spoke most frequently concerning
bills related to caring and communication across three terms (32 percent. 27 percent. and
49 percent of the total floor meetings speeches in each term. respectively). Male
legislators spoke most frequently on economic and development issues (29 percent) in the
first term; on political and government organization issues (35 percent) in the second
term. and on matters of the judiciary and social order (29 percent of the total male
speeches made) in the third term. From the perspective of particular policy areas, it can
be seen that. in the first term. 31 percent of female speeches on the floor concerned bills
related to the caring and communication area. compared to 20 percent of male speeches.
In the second term. 27 percent of women’s speeches compared to 20 percent of men’s
SPeeches were made in that policy domain. F inally. in the third term. the gap is even

larger. with 49 percent of women’s speeches concerned with this policy domain as

123

Table 4-8:

Numbers and proportions of Yuan speeches in each policy area

1” term 2"d term 3'rd term
Female Male Female Male Female Male

 

Economic
Development 102(21%) 1239(29%) 113(20%) 815(25%) 12(5%) l38(11%)

 

Defense and

 

 

 

Foreign 4(.8%) 104(2%) 53(10%) 116(4%) 13(6%) 95(7%)
Politics and
Govern 66(140/6‘) 577(14%) 144(26%) 1121(35%26(11%) 309(23%)
Judicial and
Order 61(l2%) 757(18%) 56(10%) 258(8%) 47(20%) 388(29%)
Caring and

Communication 154(32%) 857(20%) 149(27%) 633(200/6)115(49%) 329(25%)

 

Feminist
0(%) 0(%) 0(%) 0(%) l 2( 5%) 9(.7%)

 

Group
101(21%) 743(17%) 38(7%) 29l(9%) 8(3%) 52(4%)

 

Total
488 4277 553 3234 233 1320

 

Cell entries are the numbers of speeches made in Yuan meetings in each policy area.
Parentheses in parentheses indicate the ratio of that type of bills among all policy area
within each gender group

Source: the index system of the legislative participation (primary source) and the
Legislative Gazette

opposed to only 25 percent ofmcn’s speeches.

It is also to be noted that no feminist bill was put forward in the Yuan meetings prior
to the beginning of the third term. In the area of feminist bills, and as observed earlier.
only one piece of feminist legislation was put forward in the committee meetings in the
first term. That particular bill did not reach any agreement at the committee stage
and.consequent1y. no bill was sent to the Yuan meeting. In the second term, most of the
feminist bills were fiercely debated in the committee stages. Only in the third term. were
feminist bills sent to the Yuan meetings for second and third readings.

To examine the effect of gender on the number of speeches made concerning the
areas of caring and communications and feminist issues at this final stage, OLS
regressions are again applied. The unit of analysis is once again the individual legislator.
The factors affecting the frequency of the individual speech on the Yuan are generally the
same - party. age, education. constituency urbanization, and gender - as before. The
dependent variables here are the frequency of Yuan speeches concerning caring and
communication issues and feminist issues. The results are shown in Table 4-9.

The correlation between gender and floor speeches upon caring and communication
issues was positive and significant in the second and third terms. In the first term. the
model has shown that the frequency of speaking in the Yuan meeting was affected by the
level of education and party affiliation - better-educated legislators and DPP members
paid more attention and spoke more often in the caring and communication policy area.
In the second term. though the education level and party ID of each legislator was still
sufficient to explain the differences in the number of the speeches made, gender was also
Significant. By the third term. though. education and the DPP were no longer significant
factors; instead. gender and the NP now appeared to be most important in explaining the
differences. In addition. gender and the frequency of speech upon feminist policy area

have been positively correlated and the relationship is significant by the third term.

OLS models of Yuan speeches

Table 4-9.

 

1" term

Care/Comm Feminist

2"d term

Care/Comm Feminist

3'd term

Care/Comm Feminist

 

Independent

variables
Constant

Age

Education

(3ender

Rural

Party-DPP

Party-N P
Indep.

Model

diagnostics

Adjusted
R-square
F
N

-5039
(7.072)
-1.856
(1.466)
5.166***
(1.298)
5.628
(3.675)
3.166
(11.582)
12.476***
(3.081)

1.264
(4.604)
.223

5.641***
98

98

-3323
(3.166)
.161
(.680)
1.718**
(.578)
4347*
(1.798)
.845
(5.790)
3.717**
(1.209)
1.166
(2.53)
.709
(3.470)

.097

3.327**
155

1.060
(2.159)
-299
(.486)
.512
(.380)
2410*
(1.085)
-5.385
(4.424)
1.352
(.811)
4.281***
(1.219)
-2210
(3.345)

.141

4.708***
159

.042
(.184)
-.013
(.041)
.021
(.032)
.467***
(.093)
-.389
(.377)
.078
(.069)
.023
(.104)
-.083
(.285)

.135

4.526***
159

 

Note that the President and the Vice-President of the Yuan are not included in the sample

N
*p<=.05

Source: 1. The index system of the legislative participation.
3. The Electoral Bulletin of Central Election Commission, 1989. 1992. 1995

**p<=.01

***p<=.001

2. The Legislative Gazette.

Summary: gender difference in the legislative participation

At an aggregate level. I have compared female and male primary policy interests at
each legislative stage: namely. bill initiation. committee chairship. committee speech. and
floor speech. I have shown that women and men have different policy preferences with
respect to allocating their time and effort both within the groups and between groups. I

summarize the different forms of legislative participation below.

Gender difference in the policy preferences
Throughout the legislative processes I observed, women and men have different
preferences with respect to various policy domains. I summarize the top policy priorities

within the female and male groups for each legislative measurement across three terms in

table 4-10.

Tablez4—10. First policy preferences in different stages of the legislative process.

 

 

 

 

 

 

 

 

 

1" term 2nd term 3rd term

Initiation

Female Care and Com. (47%) Economics (25%) Care and Com.(42%)

Male P01. and Gov. (37%) P01. And Gov.(32%) Economic(36%)
Chairships

Female Care and Com. (25%) Care and Com.(43%) Care and Com.(38%)

Male Economics (25%) Economics (34%) Economics (36%)
Committee speech

Female Care and Com.(37%) Care and Com.(56%) Care and Com.(24%)

Male Economics (30%) Economics (29%) Economics (32%)
Yuan speech

Female Care and Com. (32%) Care and Com. (27%) Care and Com.(49%)

Male Economics (29%) P01. and Gov. (35%) Order ( 29%)

 

Numbers in parentheses indicate the ratio of the participation of that policy area among
all the policy areas within the gender groups.

Table 4-10 shows that within the two gender groups, with a few exceptions

(concerning initiation in the second term). women deputies tended to consider caring and

 

CE

1

na~

T051515) .

C11

communication issues their primary policy concern. whereas men placed matters of
economic and development policy in this position. Therefore. my findings thus far
confirm that female legislators do indeed tend to emphasize “women’s issues”. whereas

male legislators address “men’s issues”.

Gender differences in caring and communication issues

At each of the above legislative stages. I have also tried to compare the gender
difference as regards certain policy areas: namely, matters of caring and communications.
and feminist issues. I summarize gender differences for different legislative stages and
terms by deducting the percentage of male legislators’ participation in the policy area of
caring and communication from that of female legislators. Table 4-11 shows the gender

gap in this policy area that results from this calculation.

Table 4-1 1: Gender difference on the percentage of legislators’ participation in the caring
and communication issues.

 

 

 

 

 

 

 

 

 

lSt term 2nd term 3rd term
Initiation 34% 5% 25%
Chairship 7% 22% 26%
Committee speech 23% 30% l 1%
Yuan speech 12% 7% 24%

 

 

The differences between the two groups are obvious and the direction is always
positive, which confirms that women appear to put more emphasis on these particular
policy areas than men do.” The individual-level analyses of OLS regression also show
that women are positively correlated with the level of participation in the caring and
communication area and that, except in the first term, the coefficients are significant. In
general. comparing the efforts made by women and men, women as a group tended to

Place more weight on the areas of caring and communication (children. welfare.

 

19 ' . , . .

Likewise. the Slgn ofthe same measurements between women and men 1n econonnc development and
POIItICal issues is always negative. which means constantly men appeared to place more weight on these
139119? areas than women did.

education. and health).

Qender difference in feminist issues

1 also discerned the gender difference in feminist issues across different legislative
stages and terms by deducting the percentage of male legislators’ participation in this
policy area from that of female legislators. Table 4-12 shows the gender gap in this

particular area.

Table 4-12: Gender difference on the percentage of legislators’ participation in feminist
issues.

 

 

 

 

 

 

1” term 2"d term 3rd term
Initiation 16% (4.2%) 13%(4. l %) 12%(2.2%)
C hairship 2%(0.4%) 4%(2.2%) 9%(2.4%)
Committee speech 0.2998%(0.1°/6) 3.5%(1.9%) 7.1% (1.9%)
Yuan speech --- --- 4.3%(1.4%)

 

 

 

 

 

Numbers in parentheses indicated the percentage of feminist issues of all policy issues

Once again. the differences between groups are obvious and the direction is always
positive. meaning that women appear to put more emphasis on feminist policy areas than
men do. The individual-level analyses of OLS regression also show that gender is a
significant factor in explaining the gender differences in the participation in feminist
issues after controlling for other factors across the three terms (except in terms of
chairship in the second term. and as explained earlier).

In the area of feminist policy. the degree of gender difference in each term decreased
as we moved from the earliest stages of agenda setting to the latter stages. In the first
term, there was a 16 percent gender gap as regards the level of bill initiation. a 2 percent
gap in committee chairship. and a less than 1 percent gap in terms of committee speeches.
In the second term, there was a 13 percent gender difference in terms of bill initiation. a 4
Percent gap in terms of committee chairship, and a 3.5 percent gap at the committee
speech stage. In both the first and second terms. there were in essence no speeches made

at the final Yuan stage concerning feminist bills and, therefore. no difference can be

detected. However, in the third term, there was a 12 percent gender gap as regards bill
initiation, a 9 percent gap as regards committee chairship, a 7 percent gap as regards
committee speeches. and a 4 percent (approximately) gap in the number of final speeches
made on the floor. C learly. across three terms. the gender gap decreases as we move
toward the later stages of the legislative process.

In general, therefore, a funnel-shaped gender difference as regards policy
participation in the area of feminist issues can be discerned in each term. Such a pattern
of decreasing gender difference in later legislative stages has not been found in other
policy areas. This pattern suggests that gender difference is likely to be most apparent at
the bill-initiation stage, followed by the chairship stage, the committee speeches stage.
and the floor speeches stage, in that order. This appears to confirm Tamerius’ assertion
that female leadership in feminist issues is likely to be more apparent at the agenda
setting stage of the legislative process than in later legislative deliberation or other
processes. She attributed this situation to the different level of efforts. commitments, and
expertise demanded and required by different stages of the legislative process. Tamerius
(1993) found that the largest sex difference occurred in the area of sponsoring bills. the
legislative activity that required the most awareness. support, commitment, and expertise.
The more purposive involvement that an activity required. the larger was the size of the
gender gap as regards the performance of that activity (Burrell 1996).

Tarnerius’s assertion of a decreasing gender difference in passive legislative
measures remains true in my findings. In addition, though. my study across three
legislative terms also reveals that feminist issues have provoked a higher level of
legislative participation as time has passed. The percentage of legislative activity devoted
to feminist issues (figures in parentheses), in terms of chairship, committee speeches, and
Yuan speeches, increased in the third term. Comparing the percentage of the feminist

issues addressed at each legislative stage across three terms and we find that, except as

regards bill-initiation.30 the percentage increases from the first to the third term.
Examining each term individually. we find that. in the first term, the differences between
stages as regards the percentage of feminist issues addressed is the most obvious (from
4.2% -) 0.4%-) 0.1%). since most feminist bills were proposed but halted at the bill-
initiation and the committee-examination stages. In the second term, the difference
between the stages is still evident but has decreased (from 4.1%-)2.2%-)1.9%). since
most of the feminist bills have reached the stage of policy deliberation. yet few of them
are actually passed. Only in the third term is the percentage of the feminist issues
discussed at all the legislative stages more evenly distributed
(2.40/o-)2.4%-)1.9%-)1.4°/o).

By examining the legislative process across three terms. I have been able to explore
further the changes and constraints that the policy environment has created as regards
feminist issues. Policy constraints are the strongest in the first term and weakest in the
third term of the Legislative Yuan. Put another way. the policy opportunity window for
feminist issues is far more open in the third term than it has been in previous legislative
terms. Therefore. with relatively few policy constraints present, the degree to which
feminist issues are addressed increases as the legislative process moves on. Therefore. I
argue that policy environment affects the sizes and results. to a considerable degree,
displayed in any examination of legislative participation upon feminist issues. As I have
observed, feminist legislation fared relatively well in the third term compared to the first
and second terms. So what changes have occurred in the third term Legislative Yuan that

facilitate this finding? My interview data will tell that story next.

 

2” Sponsorship in Taiwan’s Legislative Yuan is continuing. that is. the bill proposed in previous terms can
be PU! in the agenda ofthe next term. Therefore. if the bills in the previous legislative term have not
reached any conclusion or did not pass. deputies can continuously bring out the old bills of previous terms
bypassing the sponsorship process. Many feminist bills were proposed in earlier terms and the discussion
and Passage was postponed till the third term.

131

CHAPTER 5
INTERVIEWS WITH LEGISLATORS:
AN ATTITUDINAL ASSESSMENT

In the previous chapter. objective measurements were used to examine whether a
gender pattern of legislative participation existed across the three legislative terms. In
general. the results confirm the behavioral hypotheses, namely. that gender affects the
deputies’ policy interests and choices. In this chapter. the psychological basis and
mechanism behind those behavioral gender differences will be explored. Since a
particular legislator's predisposition will form the basis of their legislative behaviors. I
will attempt to discover who and what is "important" to them. The question of who
deputies regard as their primary concern is related to their conceptualization of

constituency. and the question of what deputies regard as important is related to their

evaluation of policy priorities.

I conducted these interviews in order to examine whether female deputies perceive
their constituencies differently from their male colleagues (hypothesis 3), and whether
female deputies emphasize different policy priorities to their male colleagues (hypothesis
4). In addition, I will also examine whether there is a positive association between a

legislator's perception of their constituency and the possession of policy priorities

favorable to that group (hypothesis 5).

Theoretically. gender differences could be attributed to female group affiliation and
identity. whether originating from unique sociological or psychological conditions.
Carroll (Carroll 1992) argued that female politicians, who are involved with women’s
organizations outside or inside government, will be more likely than men to develop
affil iations with women’s groups and to attempt to represent female concerns and culture.

Therefore. in an attempt to explain the gender differences discovered. I will conclude this

132

chapter with an examination of the connection between women’s networks and female

legislators.
In short. this chapter commences with a brief introduction to the background of my

interview data. This is then followed by investigations of hypotheses 3, 4, and 5.

respectively. Finally. based on the theory and interview data. I will provide the women’s

culture explanation for the gender patterns found.

The interview data

The 3’d term legislature was composed of 164 members. of which 22 were female
legislators and 141 were male. and of which 84 were KMT members. 54 DPP members.
21 NP members. and 4 were independents. Of these 164 legislators. I interviewed 40 - 21
women and 19 men. The reason why the number of male legislators is lower than the
number of female legislators is due to matched sample concerns. In my sample. I tried to
interview all the female legislators. but was refused by one. I also interviewed a matched
sample of male legislators. from the same party and sharing the same constituent districts
as their female counterparts. However. a perfect matched sample was almost
unattainable. predominantly because some districts are too small to contain sufficient
male legislators who shared with their female counterparts the same constituency. the
same party affiliation. and similar experiences. Even if a matched sample could be
found, some male legislators were unwilling to accept the interview. Hence, a slight

adj ustment or substitution in the matched characteristics is inevitable, and table 5-1

Shows the distribution of my sample with respect to party and gender.

To make sure that the sample approximates to a matched sample by gender, 1 also

Chec ked and compared the other relevant variables. such as the age. education. and the

pr eVious political experiences of the two sample groups. The results of this are found in

table 5-2.

Table 5-1. The sample distribution by party and gender

 

 

 

 

KMT DPP NP INDEPEND TOTAL
8316316 11(12) 6(6) W” 4(4) 0 21(22)"""
Male 9(72) 7(48) 3(17) 0(4) 19(141)

20(84) 13 (54) 7(21) 0(4) 40(163)

 

Figures in parentheses are the population distribution of the third—term legislators. Some
of the legislators changed their party identity while serving in the legislature, but here I
adopt their party identities at the time of election.

Table 5-2 . The sample distribution by age. seniority. and education.

 

 

Mean age Mean education % who are
level incumbents (n)
Women 45 2.7 .51(12)
Men 49 3.1 .84(16)
Total 47 2.8 .70(28)

 

In education. 1 indicates high school or less; 2 indicates bachelor’s degree; 3 indicates
master’s degree; and 4 indicates Ph.D. degree

Within the sample. the average age of female legislators is 45. and that of male
legislators is 49. This pattern of age difference is not consistent with the general
expectation. Based on past studies, it was my expectation that the female legislators
would be older than their male counterparts, because familial obligations prevent them
entering politics as early as males can. The majority of my interviewees were highly

educated. with most of the women possessing either a master’s or bachelor’s degree, and
the men having earned a Ph.D.. a master’s, or a bachelor’s degree. Besides being highly
Educated. they are also politically experienced. Of the 21 congresswomen interviewed.

1 2 are senior representatives. Among the remaining nine. three served in other elected
Offices (National assembly members or Provincial assembly members) before entering
the Legislative Yuan. The other three were either previous party cadres or public servants
i 11 the executive branch. Only three out of the nine female newcomers did not have
preVious political experience in public office. Among the 19 congressmen interviewed.

l 6 Were senior members and three were newcomers to the Legislative Yuan. two of who

134

had never held any political office before entering the Yuan. As a whole. most of my
interviewees. both male and female. are experienced politicians and. therefore. as regards
the backgrounds of the two sample groups. they are generally compatible in terms of age.
political seniority. and education.

In the interview itself. each legislator was asked two main sets ofquestions in order
to test two hypotheses. The interviews lasted anywhere from 20 minutes to two hours.
depending on legislators’ schedules, interests. and degree of cooperation. When time
permitted. I asked some additional questions. but these were not intended to assist with
the systematic analysis. However, those additional questions and answers are helpful to
us in so far as they provide us with a more in-depth understanding of how the
representatives legislate. Most of the interviews were conducted in legislators’ offices, or
in small conference rooms in the Legislative Yuan, and the interviews took place between

September 1997 and January 1998. during the fourth section of the third term Legislative
meeting.
Hypothesis 3: Legislators’ Constituency Perception:
Standing for all or groups? Which groups?

In the theoretical portion of this study. I have elaborated on the dispute between
liberal individualism theorists and the identity group theorists regarding their perceptions
of representation. The liberal theorists maintain that our representative bodies should
transcend interests based on group identity, even though they may agree that different
identity groups (e.g.. women. or ethnic minority groups) have different interests. In
Contrast, the supporters of the politics of identity claim that representatives should
represent different identity groups. and particularly those that are oppressed (Young
1 990). These opposing viewpoints have aroused heated debates. both theoretical and

empirical in nature.

In order to understand legislator's relationships with their constituencies, F enno

(F enno 1978) was the first to distinguish the four types of constituencies that the
legislators can perceive of — namely. the geographical constituency. the re-election
constituency, the primary constituency. and the personal constituency. Huang (1994)
regarded F enno’s framework as applicable to Taiwan’s case. applying it to the study of
constituency service in Taiwan’s legislative districts.

Even though this is an appropriate and comprehensive frame of reference with which
to understand the legislators’ perceptions of their constituencies. it does not address
legislators’ relationship to the group constituency of my theoretical and hypothetical
interest. In hypothesis 3. I assume that women legislators are more likely to identify with
the female constituency and possess a personal commitment to represent women as a
group. Given that I am primarily interested in the constituency groups related to
women’s experience and identity. a different conceptualization scheme is. therefore,
required to capture the dynamic of representation of this identity group.

Thus. to test hypothesis 3. rather than ask a totally open-ended question about their
constituency. I asked: "As a representative. you represent many different constituency
groups. Do you consider any group’s concerns as very important to you and should be
addressed by you? For example. the business constituency, military group. civil servants.
education groups. labor groups. women. the elderly. children. or a single issue group
(6. g.. environment group)?”

In my question, I categorize the constituencies into different groups. consistent with
Thy theoretical hypothesis of identity difference. However. it is expected that different
COnceptions beyond my own classifications would appear in their answers. Legislators
haVe their own definition ofconstituency. which can be determined by region, by
oCCUpation. by social status (classes). or by certain other characteristics. Their own
chal‘acterization is also helpful for a complete understanding of the meaning of
"Constituency". Therefore. before exploring the gender differences related to the

I 68 iSIators’ constituency perceptions. I first attempt to clarify the general factors that

136

affect their views on the concept ofgroup constituency.

Factors influencing legislators’ perception of constituency

Most legislators referred to different factors that affect how they view their
respective constituencies. Even when answering the same question, respondents came up
with different ideas ofconstituency. Thus, it is clear that representation means different
things to different people. Various factors have been identified to affect legislators’
constituency perception. including electoral rules. district demography, seniority.
ideology. professional backgrounds. and gender.

First. electoral rules affect the relationship between representatives and their
constituency. Before 1991. in Taiwan’s legislature election. voters had to choose to
register between one of two rosters. territorial or occupational. before they could cast
their votes. Voters had to decide whether to register under the geographical district
roster. and thus be able to vote for representatives within their own resident districts; or to
register under the occupational roster. and vote for their occupational group
representatives. including the business. industrial. labor, farming. fishing, and educational
groups. Besides the occupational groups, another type of group representation was that
represented by the female reserved-seats system. However, unlike occupation-based

group representation, there is no independent voter roster for this system, and instead the
reserved seats are elected within the geographical district. Though the gender quotas
rernain in place. since the second-term national legislative election none of the female
I'epresentatives have been elected through this system.

Under the old occupational group system. the representatives had a clear
Corresponding constituency group from which and by which they were elected. This
Sy Stem was most applicable to a cooperative society. within which sectional
representatives were expected to represent economic-based or class-based interests.

1’10 Wever. under the authoritarian one-party regime. the system of occupation-based

representation was regarded as the KMT’s control mechanism, used to co-opt
occupational or class interests. Only members of lawful associations, who met the
criteria stipulated by law. were eligible to elect their representatives. The KMT
controlled and demobilized all modern social sectors through the preemptive
incorporation of various occupational groups (Chu 1994). Therefore. under the KMT’s
control. the group representatives reflected the KMT’s social agenda instead of their
actual members’ interests. After 1991. with newly-formed parties. the system of
occupational group representatives was replaced by party-list proportional representation,
within which each party wins seats proportional to its share of the vote in districts.
In terms of the relationship with the constituency, it is expected that
representatives elected from the party-list will have to endure little pressure from the
local constituencies. Rather. they will be concerned more with the party and group
interest. Among the 40 third-term legislators I interviewed, five were party-list members
from the KMT and the DPP. As one DPP party-list representative (interview: 98039)
pointed out, because she was elected from the party list. she does not have a great deal of
regional or local constituency services to undertake. She claims that a certain division of
labor exists. under which city councilors take over most of the local individual requests
and legislators deal with issues at the general policy level. If she receives requests related
to a local or city matter. she will pass them on to the city councilors. Generally, she
1‘ eports that, as a party-list representative, she is more often in contact with groups than
individual constituents. Because she is elected from the minority group of the party-list."
She regards herself as representing the interests of culture-based groups. educational
groups. social welfare groups (senior and children related to minority). and women.
Another DPP party-list member (interview: 98040) shares a similar view regarding his
relative freedom from constituency service. Because the local electorate does not elect
\’
2’ DPP divided its party-list representatives into three teams: politics, scholar. and minority groups.

 

138

him. he claims he is in a better position to represent minority and environmental groups.
However. there is some variety amongst party-list representatives. In my sample.
three interviewees are KMT party-list members and. among them. two were once labor-
group representatives and one is from the nursing association. The representative from
the nursing association (interview: 98036) makes clear that she identifies with the nursing
groups and that her friends in nationwide nursing associations offer her substantial
suggestions and warm support in the legislative process. The other representative
(interview: 98038) identifies various groups as being those that she is most concerned
about. including labor. military. civil servants. and women. Still another senior KMT
party-list member (interview: 98037). once the occupational labor-group representative
and now from the party-list. denies the idea of any group constituency and emphasizes

the generalization of representation. He argues that.

In the past. we used to have professional group representation. because the minority
lacked the ability of political participation and to reflect their own opinion. It has
been revoked today. Now. the national educational level is more balanced and
widespread. Group quota is still emphasized by some people. but in fact. all
politicians want to be more popular and take care of everything. Besides. our
government already divides the work into various administrative units - such as the
nuclear. technology to take good care of every profession and citizen from
different perspectives (interview: 98037)

During the interview. he repeatedly emphasized the importance of cooperating

Vvith the KMT government. He considers group representation as unnecessary since
“society now is civilized. open. and democratic, and citizens should learn to cooperate
“with government”. Ilis stand could be attributed to his senior status within the KMT as a
Party cadre and party-list representative for several terms. Generally speaking. among the
party-list members interviewed, the variety present in their answers was largely
dependent on the different backgrounds from which they come. For example, the senior
party cadre disapproves of any group constituency except when it conforms to the will of

the party; while others. perhaps indirectly elected thanks to the support of a particular

group or groups. will emphasize their connections with those groups. Most of the party-
list representatives. both KMT and DPP members, pay little attention to the individual
constituent services since no local electorate is responsible for electing them. Instead,
most of them make contact with groups and thus do identify themselves with a given
group constituency.
In addition to the party-list representatives. a majority of representatives are
elected from the geographic districts under the SNTV rule as described in Chapter 2.
Under the SNTV system. it is usually the case that more than one seat in each district is
open. which induces parties to nominate more than one candidate in each district.
Therefore. candidates must compete with competitors from different parties as well as
those from their own party. Given this. it is very important for them to build a personal
source of support rather than rely on the party to do so for them. Consequently, the role
of the faction has long been seen as an important informal mechanism through which
legislators can mobilize votes.

Since the majority of the legislators are elected from the districts, the district
demography is also a determinant factor of constituency perception amongst legislators.
Some legislators claim that they represent all the people in their districts, some legislators
from rural counties place the rural group’s interest as most important, and still other
legislators elected from the city districts highlight the middle-class interest among their
concerns. For example, the so-called old-farmer legislators, who are elected from
agricultural districts. have long been known for fighting for the interests of farmer’s.
Some legislators from city areas claim that the homogeneity of their districts has blurred
the boundary of classes. and that they therefore do not favor any particular group or class.
When they were pushed to make a distinction between groups. some identified with the
interests of the middle-class. i.e.. the majority of the city residents.

Moreover. several legislators link their self-identified group constituency with their

pr eVious professions and social status. For example. some legislators identify with

140

teachers. the military. or the minorities. because they themselves were once members of
such groups. Their former experiences or professional training has granted them the
familiarity and expertise necessary to deal with the issues related to those groups.

Some mention the difficulties that they face in attempting to distinguish between the
group constituencies because of their seniority. One female legislator (interview: 9801 I)
told me that such a question should only be asked of newcomers. because they have a
clearer idea ofexactly who they represent. She described herself as having been much
surer about whom she represented when she was first elected as a legislator. However.
over the years. since she has been re-elected many times. and she makes contact with
more and more people. such clarity has faded. Another female legislator (interview:
98003) also mentions that she relates to a number of different constituencies because of
her seniority.

By and large. given the range of factors that affect legislators’ perceptions of
constituency. a clear underlying dimension (profession, class, or gender) which can
define every legislator’s constituency clearly doesn't exist. A mixed concept of
constituency is most likely to be the case since most legislators possess different and
overlapping characteristics or identities. This also sheds light on that which I discussed
earlier regarding the concept of representation. As I discussed in the theoretical portion
of this study, a general. cohesive. and fixed gender interest transcending classes or other
identities does not exist. From the legislators’ responses to the various group

Constituents. the diversified and multi-dimensional character of the concept of
representation is re-affirmed — i.e.. even legislators who share similar characteristics or
identities might identify with some very different groups. For legislators with multiple

idehtities. exactly which characteristic or identity will prevail will also vary.

141

Gender pattern in constituency perception

Consistent with the fact that various dimensions influence legislators’ constituency
perceptions. the majority of the respondents identified with more than one constituency
group. The number of groups mentioned by individual legislators varied from none to
six. The combination of constituency perceived by the legislators as important also
varied according to the individual legislator's background, experiences, seniority. and
interests. Since I am primarily interested in legislators’ perception of some relatively
disadvantaged groups and the groups of my theoretical interest. I did not attempt to
encompass all the possible constituency groups in the phrasing of my questions. As

expected, though. the respondents provided more categories than I did. I dealt with this
by sorting the constituencies identified by respondents into several larger categories: (1)
public groups. including civil servants (such as the police, and firemen). and soldiers
(veterans); ( 2) occupational—class groups. including the middle-classes. labor, farmers.
and business groups; (3) welfare minority groups; (4) educational groups; (5) children;
(6) the elderly; (7) women; (8) environmental groups; and (9) the aboriginal. Some
respondents did not identify any group: a response that 1 labeled "non-group". The
general pattern of the constituency identified by male and female respondents is shown in
table 5-3.

From the table. a gender pattern regarding the identification of group constituencies
emerges between male and female legislators. On average, women legislators identify
27 groups as their constituency. whereas men identify 1.3. Women are more likely to
identify with the group constituency than men are in most categories of each
Constituency, including the civil servant. soldier, labor. farmer, minority, education,
children. the elderly, women’s groups, and the aboriginal. Of these constituency groups.
the differences after deducting the percentage of men from that of women is positive.

Which indicates that relatively more women than men actually identified these groups as

the” Constituency.

Table 5-3.

Constituency groups perceived by legislators

 

FEMALE MALE DIFFERENCE SIG.

 

(N=21) (N=19) ( F-M)

Public Servants % yes (n) 43 (9) 21 (4) 12

Civil servant 19 (4) 1 l (2) 8 .464

Soldier 24(5) 11 (2) 13 . 281
Occup. Class % yes (n) 29 (6) 47(9) -1 8

Middle class 0(0) 16(3) -16 .061

Labor 14(3) 5 (1) 9 .355

Farmer 14(3) 11 (2) 3 .728

Businessmen 0 (0) 16(3) -16 .061
Welfare minority % yes (11) 43 (9) 26 (5) 17 .285
Education % yes (n) 43 (9) 21 (4) 22 .149
Children % yes (n) 19 (4) 0 (0) 19 .046
The Elderly % yes (n) 24 (5) 0 (0) 24 .023
Women % yes (it) 57 ('12) 0(0) 57 .000
Environment % yes (11) 0(0) 11 (2) -11 .134
Aboriginal % yes (n) 10 (2) 5 (1) 5 .620
*Non-group % yes (11) 14(3) 26(5) -12 .355

Total groups : Mean (n) 2.7 (56) 1.3 (25)

 

Cell entries are the percentages of respondents who mention the group as a perceived
constituency. Figures (n) in parentheses indicate the number of respondents who identify
with a group of that category. The number of total groups (n) equals the sum of the
number of respondents of all the categories except the non-group respondents (*
distinguishes it from other constituency groups). The final mean denotes the number of
the total groups divided by the numbers of male and female respondents respectively, that
is. the average ofthe groups identified by men and women as two groups.

143

Among these groups. it is obvious that the gender differences are most significant
in three particular groups: those ofchildren, the elderly, and women. Female legislators
are more likely to identify these groups among their high-priority constituencies than men
are. This is consistent with the theoretical expectation originating from the sexual
division of labor in society - i.e.. that caring for children and the elderly has become the
primary responsibility of women. Women bear more societal responsibility for caring for
these groups than men do and such responsibilities have evidently been transferred into
the political arena, as well. The most marked male/female division in my sample regards
the extent to which the legislators identify with the "female" constituency. Twelve out of
21 female legislators describe women as being amongst their highest-priority
constituency groups. The figure amongst men was zero.

Only in the constituency categories of the middle—class. the business community,
and the environment. is the percentage of men identifying themselves with these
categories higher than that of women. Of these three, the difference in the middle-class
and business groups are close to being significant and, as some of the legislators pointed
out to me. men do tend to have more business connections than women. However. these
two groups are relatively advantaged groups and thus bear little relevance to the study of
the identity groups of my interest.

This general pattern across all the group categories might well imply. according to
the group scheme that 1 frame in the question, that men are less group-oriented than
women are. However. another. more direct, indicator of group-orientation is the last
category in the above table: the non-group. According to the results for this category.
more men than women reject the idea of actually representing any group interest.

These non-group respondents substitute the group-specific feature of my question
with a rather generalized characterization of their constituency. For example. the three
female "non-group respondents" either substitute the idea of group constituency with that

of an ideology. such as Taiwanese independence (interview: 98025). or with a vague

144

identification with the constituency that "appreciates what she is doing for the nation"
(interview: 98022). or with an attempt to embrace the interests of all (interview: 98003).
Some male respondents make even stronger claims to transcend the group interest.
In answering my question. one male legislator (interview: 98026) described the majority
as “the silent majority”. whose interest need to be represented, and the minority as the

“focal minority”, whose interests should not be over-emphasized. He states that.

I strongly disagree with speaking for any minority group, so I never propose
amendments for any interest groups. If they come to me, I always say no. So the
frequency that I say no every day is very high. I probably irritate five hundred
people everyday. I think a good representative should learn how to say no, instead
of learning how to say yes. Because the silent majority would not come to see you.
those who come to see you are the focal minority. If under pressure you comply
with their requests. eventually you are going to hurt the silent majority. So we
should protect the interest of the majority, instead of that of the minority. not be
tempted by the applause from the minority (interview: 98026).

Not only does he despise the concept of group interests, but he also claims to move
beyond even the interests of his own district. He provided an example of how he made a
decision regarding a policy that might hurt the majority of people in his district, Taipei,
but which he regarded as good for the nation as a whole. He proudly tells me that it turns
out “none of them complain. If you are right. no one will blame you” (interview: 98026).

Similarly, another legislator stated, “I do not make the distinction between groups or
represent any particular groups”(interview: 98004). However, by making a distinction
between representing any group and representing the majority, he argues further that he
represents the majority interest. He insists that his style is “professional, neutral, rational.
objective. and beyond party.” Not only is group interest not his concern, neither is the
party’s interest. As a KMT member. he claims that he does not work for any party
because he thinks “justice is more important than parties. the institution lasts longer than
individuals.” He emphasizes that he has never spoken or worked for any individual or
group, and that the majority opinion should always come first.

Generally speaking. the negative views regarding the concept of group

145

representation among some respondents reflect a liberal point ofview that representation
means to represent everyone in the same way without making any distinction based on
group identity. Taking any group particularly seriously is not deemed to be fair or
desirable. According to the results of my interviews, the idea of group or minority
interest is not popular with a number of legislators. especially certain male ones. Either
they do not see the necessity of speaking for any particular group by claiming that no
such minority issue exists: or they argue that group issues might conflict with the
majority interest. and that it is the latter that matters most. They contend that
representatives should transcend any group’s demands and interests, and must be able to
represent the majority. Referring to the majority interest as unbiased and objective. and
the minority group's interests as favorable and subjective, is a liberal interpretation of
representation that may well account for why some legislators (mostly male) are less
group-oriented in answering the question.

On the other hand. the fact that women are more likely to identify with groups
related to women’s unique group experiences exhibits the sort of thinking that one finds
embedded in the politics of identity difference. The theorists of identity difference
contend that group identity forms the basis for special representation and political
recognition (Young 1990) (Taylor 1994). They argue against the attempts of liberals to
reduce the plurality of subjects to one universal point of view. Rather. they consider
women’s shared gender to grant them shared interests, which have been transferred into
representative politics. Nevertheless, the question of whether such group differences
have indeed been transferred into representational politics or not cannot be answered
simply by noting that female deputies tend to identify with women’s groups. Rather.
there needs to be recognition that women have different interests and form a politically

relevant group that needs and deserves to be represented. Is this the case?

146

The recognition of and willingness to represent women’s concerns

Even though the majority of female representatives identify women as amongst their
various constituency components. such identification might not necessarily lead to a
subjective commitment to represent women’s interests. In other words. women
representatives may be concerned about the female constituency. but an overall gender
difference in representational politics may not necessarily be warranted. Representatives
could mention many constituency groups and interact with various groups at the same
time without consciously engaging in representing any specific group. For
representatives to commit themselves to representing women’s interests and issues, they
must first recognize that women are a politically relevant group that shares some common
interests.

Therefore. I asked representatives about their ideas concerning “women’s interests
or women’s concerns.” The majority of the representatives in my sample do recognize
the existence of women’s issues. Almost everyone. both men and women. agree that
women’s issues do exist and have to be addressed by government. despite the fact that
they refer to different things. Most of the issues they actually refer to under this heading
were directly related to women and had been discussed in the third-term Legislative
Yuan. Only three male KMT members flatly denied the existence of women’s issues.

Granted the recognition of women’s concerns. to examine whether women
legislators are more likely than men legislators to commit themselves to representing
women’s interests. I asked legislators. “Do you think that you and your female/male
colleagues are more willing than your male/female colleagues to represent women’s
concems?” In general, both men and women legislators felt that women were more
willing to act upon women’s issues than men were. In addition to this similarity between
the male and female responses, certain differences of approach are also worth noting.

Whereas most female legislators describe how they are committed to representing

147

women’s concerns. and most of them hold a rather disdainful view of men’s relatively
passive participation in the argument: most men admit that women are more active in
representing women’s issues. and then attempt to explain exactly why they do not pay as
much attention to them as women do.

The typical answer among women respondents is that women are more willing to act
upon women’s issues than men are. One KMT female legislator's (interview: 98002)
view was that. despite the fact that most women legislators have a wide variety of
interests and choose different committees on which to serve. etc.. in general. they share a
common ground in so far as they are all concerned with women’s issues. She says that. “
as women. we must stand out and be more concerned with those issues” (interview:
98002). A DPP female legislator (interview: 98022) also reports that. generally, women
take the lead in women’s issues; e.g., women will draft women’s bills, get the bills a
committee examination. and have the procedural members arrange for those bills to gain
Yuan second readings. Another DPP female legislator described the situation in the

following terms:

All the passages of laws about women’s rights - the social security problem. the law
of sexual assault. the revisions of the law of road security, to the law of family
violence - are pushed by women. Now the progressive revisions in the property
right and custody of the children in the family chapter in the civil law, which
concerned about women’s disadvantaged positions. all are delivered by female
legislators (interview: 98039).

The majority of female legislators also remark that men do offer passive support for.
or as some women put it. do not dare to oppose, women’s issues. Generally speaking.
most women representatives regard the male commitment to women’s issues as
unreliable. In some instances. female legislators question the genuineness of male
involvement in women’s issues. noting that: “Most men do not care [about women’s
concerns] at all. A few male legislators support them. but they do not donate time to

carrying them out” (interview: 98039). In other instances, some women legislators imply

148

that men are reluctant to support women’s concerns by saying that: “Men are very
chauvinistic. but in public, they dare not oppose women’s bills now” (interview: 98022);
or “Of course men also pay attention to these bills, because they also have women voters.
However. relatively. women are more involved in these issues than men are” (interview:
98005).

From the point of view of most female legislators. male involvement in women’s
issues is sporadic. inconsistent. and unreliable, and a distrust of male cooperation
regarding women’s issues is common. One KMT female legislator mentions that male
legislators appear to be supportive. however. whenever it comes down to the vote, some
would excuse themselves to the restroom (interview: 98038). Another women legislator

also observed that:

For example, while discussing some sensitive issues, like displaying the photos of
those who illegally consume child prostitutes, most men at first agree to co-sponsor
the bills. We collected about 60 co-sponsorships in one case. However, when a
decision had to be made, there were only six women legislators left sitting for the
bill (interview: 98005).

With respect to male support of women’s issues, some attribute it to the increasing
demands of society. rather than any sudden desire by men to represent women. However.
when comparing men’s attitudes to women’s issues across time, some female respondents
indicate that. as times change. men’s attitudes toward these issues change, too. As one
senior female member (interview: 98031) observed, when women’s bills were promoted
by female legislators in the past, many male legislators openly opposed them. For
example, as she recalled. in the second-term. when examining such issues as the property
rights of married women in the family, sexual violence, or prostitution, most men
opposed the female point of view. However. she reckons that the situation has changed
and most of the male legislators support women’s issues. Even when they do not, they
stop short of actively opposing them. She attributes this change to social trends that have

imposed certain pressures upon men. And she regards the current situation to be

149

favorable as regards the promoting of women’s bills. since men have responded to the
social pressures upon them to accommodate women’s needs (interview: 98031)

Overall then. most female interviewees consistently assured me of their strong
commitment to representing women’s concerns, and indicated that they regarded male
support as doubtful. Even though almost everyone. women as well as men, agree that
women are more active than men as regards women’s issues, very few women believe
that this division of labor when it comes to promoting women’s issues is simply and
completely a matter of gender. Instead, they argue that these differences are just as likely
to be related to an individual's characteristics and ideologies. As one woman legislator

says.

On average. female legislators. because of their own roles. feel closer to those
concerns. So in terms of attending the meeting and participating. they are more
active. However. many male legislators are also concerned about that. As long as
you respect individuals. and have some concern for gender equality, you will be
concern with these issues. However. if you do not respect human beings then. even
if you are women. you will not pay any attention (interview: 98025)

On the other hand. however, this concept of a labor division along the lines of
gender regarding women’s issues is further reinforced by the interview responses of the
male legislators. The responses of most men tended to be along the lines of arguing that.
of course women are more active than men regarding women’s concerns, because that is
the nature, the role. the duty. and the responsibility of women. Several male
representatives claim that representing women’s concerns is a woman’s natural duty.
One male legislator (interview: 98029) argues explicitly that it is human nature that
everyone knows better about him/herself, and that, therefore, female legislators would be
more concerned about women’s issues because women know more about women’s
concerns. Most male legislators refer to the need to work on women’s concerns as a
responsibility of female legislators from which they cannot excuse themselves. Since

most male respondents regard the division of labor in agenda setting as natural. it does

150

not matter to them that women propose women’s issues. and that men propose men’s, as
long as men support female legislators’ initiatives. The predominant attitude of the male

legislators’ is illustrated particularly well by one KMT legislator's response:

Of course. women legislators are more actively involved in these issues [women's
issues], and they should be. And men legislators also do not think they should
compete with female legislators on this agenda. Legislators do compete in agendas.
however, not in this area. If you do. people will feel strange [he laughs]. and it is
not necessary. It is natural that the leading role [in women’s issues] is played by
women legislators. and we support them. We cooperate with each other to advance
their passage (interview: 98004).

Most male interviewees define their own role in the passage of women’s issues and
bills through the Legislative Yuan as being passively supportive. In the previous chapter.
I described legislative activity as including everything from active agenda setting to
relatively passive levels of participation and, among this variety of legislative activities.
most male legislators clearly take their passive role in matters of women’s interests for

granted.

Not speaking on these issues is a type of support. On the contrary, the more you
speak might adversely become a procedure filibuster, which increases the difficulty
and delays the passage. Most of time, signing it. raising hands and supporting it.
that is enough (interview: 98004).

In general, most male respondents explain and justify their limited support as
regards women’s concerns by revealing that they perceive a gendered division of labor in
such areas as natural. This common perception of “spontaneously active women,
indispensably passive men” regarding matters of female concern can be further reinforced
by a certain male perception that their support of women’s issues should be taken as
flattering by female representatives. For example, one DPP male legislator (interview:
98026) comments that female representatives do not need to worry about support from
men, because male legislators. “for the sake of good manners, have no reason to oppose”.
Though he does not always totally agree with some female legislators, believing that their

biases and emotions can go too far. he adds that. “as long as everything is reasonable.

151

they do not need to worry about male support. men are still quite flattering”(interview:
98026).

To sum up. in testing the third hypothesis of gender differences as regards the
concept of constituency. I have found that female legislators are more likely than their
male counterparts to include women. the elderly. and children as constituency groups
about which they care. and that this is consistent with their traditional care-giving roles.
The most significant gender difference has been that more than half of the female
representatives identify women as being amongst their constituency whereas no men do
so. I have also shown that despite a common recognition of the existence of women's
interests. female legislators feel more obligated and are more willing to represent
women’s concerns. The attitudinal differences with respect to the commitment to
represent women’s concerns is further manifested in both male and female responses to

the division of labor by gender that can be observed in the Legislative Yuan.

Hypothesis 4: legislators’ policy priority
Is women’s orientation merely symbolic or more?

Male legislators at most support your stand. but if you ask a male legislator to
actively lead [my own Italics] and carry out women’s issues. it is hard to imagine it
happening. As women. we are relatively more flexible. we can bring out a neutral
issue; for example. judicial reform. education. culture. economic, or defense.
Women can lead in these policy areas, but no men lead and bring out any women’s
issue. Therefore. the women’s issues must rely on women legislators (interview:
98028).

While asking female legislators what they perceive to be the differences between
male and female legislators’ involvement with women’s issues. one female legislator
replied with the above. She was not the only woman to point out that the main difference
between the genders lies less in whether legislators “support” the issues or not. but in
whether they will consistently “lead” the issue from the very beginnings of the legislative

process. Even when I point out to certain female legislators that, according to the public

records. some men do actually participate actively as regards women's issues, the
responses are less than enthusiastic. One female legislator (interview: 98039) felt that it
depended upon the issue. If the issue was very obviously one of injustice, domination. or
indicated a clear lagging behind the social trend that had be corrected, men will
sometimes speak so as to show their progressiveness. But she goes on to emphasize that.
if men were asked about their "priority bills", they would never mention women’s rights.
She claims that it is the priority bills that make the difference.

This suggests that in passive legislative participation. men show no significant
difference from women in supporting women’s concern, but that. as regards more
energetic legislative activity, the gender gap is somewhat larger. It is consistent with my
previous findings that the more purposive involvement that an activity requires, the larger
was the size of the gender difference pertaining to the performance of that activity.

Is it true that women and men have different policy agendas? So far, from the
interviews. we have seen that, compared to men, women are more committed to
representing women’s concerns. This commitment indicates a symbolic association
between women legislators and their female constituency. but it does not necessarily
imply that there is a gender difference at more critical levels, such as agenda setting,
which require enormous individual awareness and expertise in addition to considerable
commitment and support. Given legislators’ limited resources and institutional
constraints, they must obviously weigh up different policy interests when deciding which
agendas to pursue. Thus, legislators could well make supportive gestures toward various
causes without devoting the time and effort necessary to "lead" them all. This transition
from “supporting” to “leading” in women’s issues is a vital one that indicates whether
legislators are willing to move beyond the merely symbolic and begin to substantively
represent women. Therefore. it is worth discovering whether gender differences can be
discerned in this area. too.

Hypothesis four assumes that women and men tend to have different emphases as

153

regards their policy priorities. Because of the gender differences derived from
socialization and life experiences. women and men tend to encounter different problems.
associate with different groups, acquire different policy expertise. and have different
policy interests. Empirical studies from other countries show that female representatives
tend to give a higher priority to matters that concern women, children. and the family
than do their male colleagues (Boles 1991; Dodson 1991; Flammang 1985; Havens and
Healy. 1991: Kathlene 1989; Leader 1977; Saint-Germain 1989: Tamerius 1993; Thomas
1994)

An empirical study of Taiwan’s voters also found that female voters tended to
emphasize political stability. social domestic issues, education, and culture issues,
whereas male voters tended to emphasize the national defense, foreign relationships, and
national identity issues (Fan and Hsu 1994). However, will such gender differences
amongst voters translate into legislative politics? Are women more inclined to emphasize
different policy priorities from men in the legislative process? Or are women in
representative politics actually doing (or not doing) much the same as their male
colleagues?

To examine policy priorities amongst the legislators. I asked legislators an open-
ended question, requesting that they each indicate “the three most important bills s/he
thinks about and spends most time on.”" The 40 respondents mentioned 120 bills
between them and. whenever a respondent mentioned more than three bills as being a
personal priority of theirs, I coded only the first three mentioned. Those bills are first
classified into 19 sub-categories. and are then collapsed into the seven major categories
that I regard as important and relevant to this study. The seven categories are: (1)

economic and development issues. including economic issues, budget. finance and local

 

’3 To avoid leading their answers. I asked interviewees this general question about legislative priority
before asking their identification of constituency and their subjective attitude on participating on women’s
issues.

154

development: (2) politics and government organizations; (3) defense and foreign affairs;
(4) judicial and public order issues: (5) caring and communication-related issues.
including welfare. health. children. education. and communication; (6) feminist issues;
and (7) other group issues (labor. agriculture. environment. and ethnic minorities).23 The

results are shown in table 5-4.

Table 5-4. Policy priority listed by legislators

 

 

 

 

FEMALE MALE DIFFERNCE SIG
(N=21*3=63) (N=19*3=57) (F-M)
Econ. & Develop. % yes-(n) l4 (9) 9 (5) 5 .485
Defense & Foreign % yes (11) 10 (5) 7 (4) 3 .872
Politics & Govern. % yes (n) 2(1) 37 ('21) —35 .000
Judicial & Order % yes (11) 5 (3) 2(1) 3 .355
Caring & Commun. % yes (11) 43 (27) 35 (20) 8 .538
Women issues % yes (11) 24 (15) 0 (0) 24 .002
Other group issues % yes (11) 5 (3) l l (6) —6 .385
Total : (N) (63) ( 57)
Respondents were asked to list three policy priorities. Entries are percent of that policy

area as policy priority. Numbers (n) in parentheses indicate the frequency of the issues
being identified as policy priority.

In my sample. ten female legislators have women’s issues listed among their three
policy priorities. and three out of the ten include them more than once. An accumulated
frequency shows that the women’s issues area is one of the most popular policy priorities.
being mentioned 15 times among female legislators. However, as none of the male
legislators include this policy area among their policy priorities at all. the difference
between the sexes on this point is obviously significant. In other traditional women's
interest areas. such as care and communication, including education, welfare, health. and
children, which women mention the most (27 times), the difference is not so significant.

Interestingly. women exceed men in terms of the frequency with which they placed

development policy on their priority list. This reverses our earlier findings that women

 

2" The category varies in their coverage and specification in general policy terms; some policy areas are
broader and less specific. The category was set up based on the category 1 used in the chapter 4.

155

are less active in economic policy areas. However, this difference is insignificant. and
could have been caused by sample variation. On the other hand. the other significant
difference between sexes lies in the area of politics and government organization, which
is listed 21 times by men as their first priority. Many issues in this policy area24 are
related to constitutional. institutional, and political reform and many male legislators
consistently emphasize the importance of these structural issues and are apparently more
enthusiastic than women. This category has traditionally been viewed as men’s interests
in the public domain. This is also, of course. an area that contains a great deal of
controversy and political infighting and some of the female legislators mention in their
interviews that they prefer the domestic and social economic issues better for this very
reason. Overall though. the differences between male and female legislators as regards
policy priorities correspond. to a large extent. to the different priorities found among

women and men voters.

Difficulties in changing the agenda

Given that women and men have different policy priorities. especially in certain
policy areas. will women and men act similarly when attempting to change the political
agenda? In other words. “If more women enter the Legislative Yuan, do you think there
will be any changes in the agenda?” This question was designed to elicit the
interviewee's opinions regarding the agenda dynamics of the Legislative Yuan. and the
majority of my respondents pointed out that agenda change will only occur in the
women’s rights area. as most gender issues have been led by women in the past.

However. as a whole. it is expected that the legislative agenda will not be substantially

 

3’ The issues in this category include government organization (organization of national assembly and
police department). constitutional restructure (reorientation of provincial government and self-ruling clause
of decentralized local govemments). electoral rules (methods of electing the president and vice president).
china policy (relationship between two sides of Taiwan Strait and territory claim). political reform (public
property claim ofthe officials and compensation for political prisons under the authoritarian era),and merit
system (the recruitment and examination of public servants).

156

changed.

It is understandable that. even though most female legislators engage more actively
in women’s issues than do men. some calculations still deter them from pushing the issue
too far. One NP female legislator (interview: 98023) summarized several constraints on
women that can deter them from participating and leading in the area of women’s issues.
F irstly. the diversification of the backgrounds and characteristics of female
representatives has prevented women from forming a collective entity with which to
bring about any significant agenda changes. Secondly, women do not like to be
stereotyped as concerned only with themselves. Rather. they would like to be seen as
broad-based representatives. Thirdly, women voters are not independent, and, therefore.
it is difficult to encourage the female electorate to combine in their support for women
legislators. Because of these concerns. even though most legislators agree that women
would make some changes in the agenda as regards women’s issues, most of the
respondents step back from predicting a substantial degree of agenda change. even with
an increase in the number of female representatives. It is important to examine the
possible factors which prevent women from achieving significant agenda change in the
legislature, since these factors might constitute the gap between the subjective perception
and objective behaviors of legislators’ policy priorities. Therefore. I will now examine,
one by one. the difficulties associated with the placing of women’s issues at the top of the
list of possible agenda changes.

First. a variety ofobjective characteristics and subjective identities divide
representatives along various dimensions. Legislators’ multiple objective characteristics.
such as class. profession. and vulnerability to party influences. will affect their legislative
activities. Many legislators report that, outside of women’s rights issues, they do not
think women would make any significant difference to the general political agenda.
because most of them come into the legislature with their own professional bases, parties.

and different backgrounds still fresh in their minds. The diversification of women

157

representatives has prevented them from building any substantive entities through which
to bring about any agenda change. For example. one KMT female legislator (interview:
98002) tells me that her long-term interest in education and culture has long outweighed
her interest in women’s issues. even though education is still traditionally regarded as
being within women’s purview. She also emphasizes that many other women participate
in a wide variety of policy areas. such as the budget, defense. and interior fields.
particularly in the third term. She adds that. if the number of women legislators
increases. and everyone has different interests as they do now. the agenda would be
unlikely to change. One male legislator (interview: 98032) makes a similar point that
increasing women representatives might lead to the opposite result to that intended - i.e..
that the women would become unable to reach agreement on issues given the ever-
growing number ofdiverse experiences and backgrounds that would suddenly be present.

Many legislators are unconvinced that being a woman would necessarily change the
political agenda; rather. they think that it is background, resources. and training that will
make the difference. As one KMT male legislator (interview: 98004) points out,
“professions exert more influence than gender does. Even though they are all women
legislators. they are divided by parties.” Another KMT male legislator (interview: 98016)
also argues that whether women will succeed in changing the political agenda depends on
the backgrounds and class from which they come. He makes a general example that
fishermen are more likely to represent fishermen, and laborers represent labor; however.
if women are from an upper class. they are not going to represent women from different
classes.

In addition to the varied objective characteristics that divide them. legislators'
subjective identifications also shape their legislative orientations. The subjective
identification is sometimes more critical to the real representation than the descriptive
resemblance. The simply descriptive presence of women in the Legislative Yuan would

not lead to any agenda change. As one DPP female legislator (interview: 98022) claims.

158

the key to making any difference depends on legislators’ conscious identification with

other women. not on simply being a woman,

If a woman does not self-examine herself about women’s role in human history; that
is. does not understand that women’s rights are oppressed by a patriarchal
mechanism. she probably only has the body of a woman. without the mind of a
woman. If so. when she gets into the system of political power. she will continue to
use the same ways she learned from male power struggles. in acquiring and
maintaining her power. This is because she does not have the independent and
critical thinking ability. which is the key to making any difference. However,
women’s consciousness has gradually arisen in society, and women’s votes are
getting more and more attention. therefore. those women legislators with women’s
bodies show their concerns about women’s bills. This is because she wants to share
those votes, which is in fact viewing women as voting machines. It is women
representatives without a woman's consciousness (interview: 98022).

Similarly. one DPP male legislator (interview: 98026) argues that the increasing
presence of female representatives will not make any difference if the majority of women

deputies are not independent political agents. He comments that.

I do not think there is going to be any difference with more women coming into the
Legislative Yuan. If we were to examine carefully where most women legislators
came from in the DPP. as well as in the KMT, of course, some of them participated
in earlier opposition movements for democratization. however, most of them were
from family politics, being someone’s daughter or wife. If we do not take these
women into account. then the proportion of women politicians could be even less.

On the contrary. he argues that men could substantially represent women without
descriptively resembling them. Men could be good representatives of women, as a non-
labor member could be a good representative of labor, as long as men identify with
women’s concerns and possess adequate expertise. Even though men are not
descriptively women. they could subjectively identify with women. And once men
identify so. they could be better representatives for women’s concerns than are

unqualified women. He adds that.

As to the question of whether we need women representatives to get bills that

. advantage women. the answer is yes and no. For example, in the sexual assault bill.
women pushed much harder than men did. However. I think for bills related to
labor. we do not need labor representatives to make them. For example. in the US.

159

those representatives who represent labor are not laborers; they are probably
lawyers. who place protecting labor’s interest as their main concerns. Today, a male
lawyer who orients himself as the guardian of women interests might not be worse
than a female legislator. Because he is more professional and efficient. unlike an
unsophisticated woman legislator [Ms A], even though she is a woman, she doesn’t
know how to promote them [the bills] (interview: 98026).

As a result, increasing the number of female legislators might not necessarily change
the agenda. because it depends on various objective identities and the subjective
consciousness of the representatives.

Secondly. other subjective obstacles also obstruct women’s willingness to bring
about significant agenda change. A prominent consideration is the fear of being labeled
as "narrow.” For example. one NP female legislator (interview: 98003) comments that
women in the legislature should not only choose a "woman's" agenda, or “it will make
people think we have little noses. little eyes. and are too narrow” (interview: 98003).
Although well known for her concerns for women’s rights issues. she does not like
herself to be confined to these issues. She suggests that some women legislators only
focus on women issues because of the votes involved, which for her is too narrow an
outlook. She argues that female legislators should broaden their vision, looking at certain
problems from a higher and broader angle.

Another DPP female legislator (interview: 98022) also dislikes the fact that many
women are stereotyped as belonging to the education and interior committees. She states
that “not only good boys should go to the military, good girls should be able to do so.
too” (interview: 98022). Therefore. in order to challenge these female stereotypes. she
entered the defense committee to let the public see women at work in such an area. She
also points out that many female legislators confine themselves to the education
committee. which she regards as unnecessary. By consciously avoiding the stereotype
imposed on women legislators. some women legislators can make their way to some
nontraditional women’s fields and thereby avoid assuming a role in the processes of

female agenda setting.

160

Such a fear of being labeled as narrow is not only found among female legislators.
but also gains reinforcement from male legislators. One New Party male member
(interview: 98008) also mentions that women should not concentrate on social security.
violence. and family issues. yet neglect China and foreign policy. He claims that
legislators should be more comprehensive, not only concerned with taking care of
students or children. but also willing to do whatever their district electorate ask them to
do.

In addition to the issue of narrowness. a perception of insignificance and non-
salience is also often associated with women’s issues. Some male legislators do express
their feelings that women’s issues deserve little attention. It is regarded as fact that
women tend to focus on soft issues. however, as one male legislator (interview: 98004)
states. it is only productive on very limited subjects. He argues that it does not matter
that in this type of agenda only a few people are concerned and participate, since it
involves little opposition and controversy, and once the bills get proposed, they are
passed easily. He makes the example that in gender equality issues. men do not oppose
them. and that’s enough. Another male legislator (interview: 98040) shares the same
view that his limited participation in those issues is because he considers that other issues

are more important for society as a whole.

On women’s issues. most of the time. women propose. and men support. The
national policy is wide-ranging. and complicated, some legislators [women
legislators] would find the issues that she finds familiar, easy to handle and gain
attention. Every time, some accidents happen, like the Peng. Wan-Ru’s [a woman
activist] event. they [women] can use it as an agenda to raise attention, unlike me,
because I do not use those media issues. I do what they are not familiar with and it
is more important for society. So in women’s issues, it is not the case that women
are more concerned. I am less (interview: 98040).

He contends that it is not the case that he is not concerned with women’s issues,
rather that other more important issues demand his attention and efforts. Both of the

male respondents argue that it is unnecessary to engage actively on women’s issues and

161

give them a degree of priority, based on the judgement that women’s issues are less
controversial. less oppressive. less substantial, and less important.

Thirdly, the perception of the inactiveness of the female electorate also constrains
the prospect of women’s issues getting onto the primary agenda of the Legislative Yuan.
Some respondents point out that. to bring about significant agenda change. women
constituencies will first have to become more active. before women in the Legislative
Yuan can make any difference. One common example is raised by more than one
legislator. both male and female, to demonstrate the lack of subjectivity of women voters.
They take the example that ifa handsome male candidate competes with a female
candidate. the female electorate may still vote for the former instead of the latter (e.g..
Mr. Ma. the newly elected Taipei mayor is very popular with women voters because of
his handsome appearance). Using this example, a male legislator (interview: 98004)
laughs and says that the faces and images of male candidates sometimes easily cheat
women voters. He ridicules the irrationality of women voters by stating that “Women are
honest to their feelings. What they like. they choose”(interview: 98004). A female
legislator (interview: 98023) mentions the same example and regards this kind of
electoral culture as a discouragement to independent and autonomous women who may
wish to participate in politics. This perception of the passiveness of women voters has
been seen as related to the relatively low profile of women’s issues in the Legislative
Yuan. As one DPP senior male legislator (interview: 98040) states:

If women's political participation was more active. women’s issues would get more

attention. If women vote more actively, even men representatives will consider their

issues important for votes’ sake. The general idea that the voting behavior of
women in Taiwan’s election is more passive and obedient to parents and husbands
affects the fact that women’s issues are getting less attention. But this situation is
gradually changing. since the number of independent women voters is increasing.

Women’s political participation in terms of voting will affect women’s issues’

policy direction. I think it is even more important than increasing the number of

female legislators. There would be no difference because all are for votes
(interview: 98040).

As I described in Chapter 2. the political participation of women on a large scale has
been increasing along with the development of democratization and socioeconomic
modernization in Taiwan in recent years (Chou, Clark. and Clark 1990). With an
increase in the quality and availability of education, of labor participation, and profession
choice. as well as democratization itself. the gender gap has decreased in the past ten
years. However, many legislators still doubt the subjectiveness of women voters in
pursuing their own agendas and their degree of support for female candidates.

To sum up. if we are to compare the answers as to whether future agenda change is
guaranteed by introducing more women into the politics with those to the earlier question
of whether women are more willing to represent women’s concerns than men, the
answers would appear contradictory. While the answers to the second question would
appear. generally. to be in the affirmative, responses to the first one are far more
problematic and ambiguous. While most believe that the only possible agenda change
would occur in women’s issues. some of them do express the opinion that agenda change
depends on many different factors. and they also question the willingness of women in to
actually bring about substantial change. In short, this section of the paper raises the
possibility that women will not necessarily represent women just because they share the
same sex. This is because the variety of considerations that divide women legislators
may well prevent them from bringing forth any significant changes in the Legislative

Yuan.

Hypothesis 5: The association among different dimensions of gender difference
Even with the constraints against projecting a significant agenda change. all in all.
significant gender differences with respect to legislative participation (hypothesis two).
constituency perception (hypothesis three), and policy priority (hypothesis four) are
found to be present in Taiwan’s third term legislature. The question that remains to be

answered is whether a consistent and positive link exists between these dimensions of

163

gender difference, or is the gender pattern actually random.

1 first checked for consistency between the behavioral and attitudinal measures of
policy representation.” in other words. for an association between hypotheses three and
four. In the previous chapter. I examined the actual behavioral difference in the different
legislative stages. from agenda setting to policy deliberation (hypothesis two). Earlier in
this chapter. by interviewing legislators. I asked respondents to report the three policies to
which they attach the most importance (hypothesis three). The former provides me with
an objective measure of policy representation. whereas the latter generates a subjective
measure. Examined together. in most of the issue areas (caring and communication.
women, economics and development, and politics and organization) a significant and
positive correlation is present between the subjective identification of policy priorities
and actual behaviors in agenda setting (especially as regards bill-initiation). Only in two
policy areas - those of judicial matters and defense and foreign affairs - was there no
significant relationship identified between the subjective identification of policy priorities
and the objective measures. However, the frequency of these two issue areas being listed
as policy priorities is too small to extrapolate any pattern from. Overall then, the pattern
of reported policy priorities is consistent with that of the actual policy behavior.

Next, I turned to examine the relationship between constituency perception
(hypothesis three) and policy priority (hypothesis four). As discussed earlier. I asked
female legislators “do you think you and your female colleagues are more willing than
your male colleagues to represent women’s concerns? Why? Or, alternatively, why not?”
From that question. I expected to elicit their subjective willingness and commitment to
representing women as a group. However. many female respondents often spent more

time addressing the observed actual gender differences as regards participation in

 

2" My data set of the legislative participation covers from the 85'h session of the first term to the third term
(1990-1997). Here. I only apply the third-term legislators’ actual legislative participation from the data set.
since the corresponding interview data on legislators’ self-reported of policy priority is only available for
the third term.

164

women’s issues, since the answer to the above appears to them to be both in the
affirrnative and self-evident?" Their answers often ended up simply demonstrating to me
how they actually perform the role of furthering female concerns. In answering the
question regarding the subjective commitment to representing women’s concerns, some
female legislators spontaneously suggested that the gender difference will be more
prominent in the area of policy priorities. because women would be more likely to list
women’s issues as a policy priority, whereas men will not. Many respondents have
directly linked two different questions together: the question of a subjective commitment
to representing women. and the question of listing women’s issues as their policy priority.
And many respondents imply a consequential link between these two; that is. the
subjective willingness to represent one particular constituency (women) should lead to
the actual policy representation of that constituency.

From the interview data I have analyzed so far, there appears to be a corresponding
pattern of gender difference between the act of identifying women as a constituency and
the listing of women's issues as policy priority. In general, women tend to be more likely
to identify with the women’s constituency and to identify women’s issues among their
policy priorities. Of21 women legislators interviewed, 12 mention women as their
constituency and 10 report women’s issues among their policy priorities. Conversely. no
men either identify women as a constituency or report women’s issues among their top
three policy priorities. In addition. the men’s own policy priority is the area of politics
and government organization. whereas for women it is the caring and communication
issue. In constituency identification, women are more likely to identify with children and
the elderly than men are.

Hypothesis five postulates that a positive association between legislators’

 

2" Their answers to the question why women are more willing to represent women’s concern are as
straightforward as those of male respondents. they regard it as the way it should be. since women know
more about these issues. and women must do so. Most respondents attribute the gender difference of the
subjective willingness to represent women to the different nature and the experiences between the genders.

165

constituency perceptions and policy priorities exists with respect to representing women’s
interests. That is. those who are more likely to identify with a female constituency and
commit to represent women’s concerns. are more likely to list women’s issues as their top
policy priority. So far. most of the patterns regarding constituency perception and policy
priority seem to fit the gender line nicely. However. legislators are also re-election driven
and have different considerations and constraints to face in actually carrying through the
implications ofthese subjective gender differences. The various factors influencing
representatives’ choices of policy priority include some personal characteristics (age.
education, and seniority), their party affiliations, and their district demography. all as
discussed earlier. To examine the causal relationship between the constituency
perception and policy priority. I must. therefore, include these factors, as well as the
identification with a female constituency. as independent variables. The dependent
variable here is whether the legislator considers women’s issues to be among his or her
policy priorities. which could be either 1, yes; or 0, no. Given the fact that the dependent
variable is a dichotomous variable with a distribution of0 and l. 1 model it in a logistic
regression as shown in table 5-5.27

The results show that whether legislators perceive women as a constituency is a
significant variable in determining whether or not they would list women’s issues among
their policy priorities. Other variables cannot explain this significant variation of the
dependent variable. In short, I have first affirmed the consistency between the actual
legislative behaviors in the agenda setting (hypothesis two) and the perceptual policy
priorities (hypothesis four) of the third-term legislators. A positive correlation between
legislators’ perceptions and behavior in agenda setting is apparent in most of the policy
areas. Furthermore, the identification with a women constituency (hypothesis three) is

causally linked to the perception of women’s issues as a policy priority (hypothesis four).

—_I_

37 . . . . .
whether women Issues as polrcy pr1or1ty=a+b l partyd+b2partyn+b3age+b4const1tuency
urbanization+b55eniority+B6women constituency

166

Therefore. which cmzstituenc)‘ legislators perceive to be important. to a large extent
decides what issues are important to them. Overall. this demonstrates a consistent gender

difference with respect to various legislative dimensions.

Table 5-5.Logistic model of legislator’s policy priority on women’s issues (yes=1; no=0)

 

 

Independent Variables B (S. E) Exp( B)
Constant -.9706 (3.124)
Party-DPP -.1075 (1.41 13) .8980
Party-NPP 1.8164 (1.7552) 6.1496
Age .4352 (.9001) 1.5453
Education -.7475 (.8079) .4736
Seniority .8420 (1.5037) 2.3209
Constituency urbanization -.33.6734 (179.256) .0000
Women constituency 4.0755**(1.5436) 58.8801
Model diagnostic
-2 log Likelihood Chi-Square 27.757
Model chi-square 21.608 ( .0030)
Correctly predicted 87.5%

 

The seniority of the legislators is a dummy variable, with l as senior. 0 as newcomers.
The constituency urbanization is measured by the agricultural population divided by the
total employed population within the districts. For those party-list and non-district
members, I substitute the measure with the national agriculture population divided by the
national employment population. The variable of women constituency is a dummy
variable, taking women as the constituency group set as 1, and not as 0. The party-DPP
and party-NP is also dummy variable, with DPP and NP members set as 1.

A "women’s culture" explanation of the gender difference

All in all. I have observed a significant gender difference as regards women’s issues.
with respect to constituency perception and policy priority. On an individual level. I also
try to forge a causal link between these two seemingly correspondent gender patterns. by
examining the positive correlation and causation between seeing women as a "most-
concerned" constituency. and listing women’s issues among top-three policy priorities.
All that is left to do is to provide a theoretical account of just why these gender gaps in
legislative perception exist.

Carroll (1992) provided a plausible explanation for the gender difference in terms of

167

identification with women’s groups and policy priorities. She states.

women’s networks and organizations. both inside and outside government. play a
vital role in linking women’s officeholders to women’s culture. Women’s
organizations and networks help to maintain and support the responsibility of
women officeholders to represent women’s shared interests and to act as an advocate
on behalfof women within the institution in which they serve (Carroll 1992, 25).

She posits that women’s networks. which links legislators with female culture.“
should affect legislators’ identification with women. too. I attempt to explore if such a
women's network with women’s culture exists in Taiwan’s Legislature Yuan and, if so.
how it relates to women’s constituency relationships and policy priorities. by examining
the connection between women legislators and women’s networks both outside and inside
government.

Outside the government, two types of women's organizations should be noted -
feminist organizations and traditional women's organizations. The traditional women’s
organizations are public-service-oriented membership organizations. which aim for
promoting traditional women’s interests. as opposed to feminist groups aiming for
promoting gender equality and women’s rights. The latter are more active in the process
of law making. The information from the roster of third-term congressional officeholders
shows that none of the women legislators are members or leaders from feminist groups,
but some of them participate informally in those groups’ activities. On the other hand.
most of the KMT women legislators either hold positions or have interaction with
traditional women’s organizations.

Different party members do have different ties with different women’s groups.

Among l 1 KMT female legislators. five held positions in local traditional women’s

 

3” Carroll defines women’s culture as “women as socially constructed. gendered beings have certain shared
interests resulting from their position in the sexual division of labor and the oppression they face even
though the strength and nature of the sexual division of labor as well as the manifestations and
circumstances of their oppression may vary considerably depending on race, class. sexual orientation. and
other cultural differences. Women’s shared interests which can lead to common values. perspectives. and
patterns of interaction. define ‘women’s culture” (Carroll 1992. 26)

168

groups, historically associated with the KMT party machine; and three held positions in
the KMT's central women's section. Among KMT women legislators. an almost perfect
overlap appears between those who belong to the traditional women’s membership group
and those who identify women as one of their constituency groups. Seven out of eight
women legislators who have established networks with traditional women’s groups
identify women as their constituency. and. among those female legislators that identify
women as their constituency. only one does not have a close tie with traditional women's
groups. On the contrary. in the roster. only one DDP woman legislator was once the
party’s women's sector’s executive; and only one NP women legislator holds the
leadership position in a professional women's group. Since the KMT has long controlled
local politics. and most of the traditional women organizations have developed in parallel
with such local politics. KMT female members have clearly outpaced other parties’
members in establishing close ties with traditional women’s groups. However. this type
of women's organization wields its electoral influence through the party machine, the
historical association between traditional women's groups and the KMT having given the
party the edge. And this. in turn. reinforces the relationship between the KMT women
legislators with those organizations.

The relationship between the KMT establishment women and traditional women’s
organizations is maintained through a formal relationship, by being a member of, or
holding a position within. one of those organizations. However, the relationship between
establishment women and feminist groups is generally a temporary, issue-oriented.
coalition. The roster shows that none of the women legislators are members of feminist
groups, or come from feminist groups. The relationships between establishment women
and feminists groups can thus be seen as dynamic and fluid. From the interview, only a
few of the women legislators mention their relationships with feminist groups. yet. the
relationship is not always seen as positive and cooperative.

The KMT women legislators’ attitudes toward feminist groups are ones of

169

considerable doubt. However, because of the political power wielded by feminist groups.
some female legislators, even if they hold different views. will choose to remain silent.
As one KMT female legislator (interview: 98002) explains, one of her reasons for not
"leading" women’s issues is because she holds a different perspective of what the female
agenda should be from feminist groups. She argues “there are different types of women's
organizations. some are radical, and some are conservative. If they are too radical, I
cannot agree with them.” From that, she mentions some cases that regard the promotion
of gender equality, ones which she fears might over-protect and obstruct women’s
development and growth. She regards some of the feminist law-making as problematic,
too. However. during our two-hour conversation. she asked for an assurance of
confidentiality several times, because she knows that some people like to use such issues
to attract votes and, if she opposes them, she will be attacked. She complains that “I do
not want to irritate others [feminist groups]. nor endorse, so I am just involved less in
the Legislative Yuan. Democracy has its defects coming along with its votes, and what is
of most concern to legislators is applause and votes, nothing else I am very rational.
However, if I do not have votes, I lose” (interview: 98002).

In the DPP. the responses of female representatives toward feminist groups are more
accepting. One female legislator (interview: 98039) has assigned one of her assistants to
hold meetings with feminist groups on a monthly basis, meetings from which the
legislator can learn more about the organization’s plans and the bills in which they are
interested. Besides these regular meetings with feminist groups. the representative told
me that. sometimes. they act with feminist groups in holding public hearings, journalist
meetings. or meetings with administrative officers. Another DPP women legislator
(interview: 98025) also participates frequently in the activities held by feminist groups,
such as various seminars and meetings. However. a new DPP woman legislator
(interview: 98022) complains that the feminist groups are not so welcoming to other

women legislators as they often come to those with whom they are already familiar. and

170

she regards this as a constraint on feminist development.

One NP female member (interview: 98028) also regards the relationship between
female legislators and feminist groups as positive and mutually supportive. She notes
that. “Women legislators and women's organizations help and support each other. They
need us to help them understand or change laws. given the legislative Yuan’s constraints.
We need them to help us understand about social trends. and their expertise and research
help our work.”

Still. another NP woman legislator (interview: 98017), who is active in the
furtherance of feminist bills. maintains a rather discontented attitude toward the
relationship between herself and the feminist groups. even though they have worked
together many times. She complains that:

In the process of working with women's organizations, I always think they are very

demanding on legislators. and it is unnecessary. They do not take legislators as

cooperating partners. rather. they take you as the demanding and monitoring
subjects. They think if they do not push and force you, you will not do it. They
even took my version of amendments [in a civil law] as theirs, but in return they
blamed me. So I do not care much about them now. what I want to do is not because
of their push. but because I feel it is an importantjob, so I do it. Like last time, we
brought back some old laws to the middle I thought if you just said “thank you
for the help and we need more efforts”. then I would continue. But if after all the
effort 1 had put forward. you still said “why do you just bring it to the middle? Why
can’t you finish it?” I am sorry. please do it on your own, I am not interested in

working with you. I dislike some of this kind of organizations. If you push
something, you should be benign ...(interview: 98017).

Even though this legislator is very active on some women’s issues. she does not
mention women as her constituency. The reason for her not identifying with women
might be attributed to her discontent with women’s groups from her past experiences; it
could also be the case that she was primarily concerned about another constituency group
(children).

In addition to the women's organizations outside of the Legislative Yuan,

establishment women have also organized caucuses within the Yuan. Two women’s

171

caucuses are important in binding congressional women as a collectivity - the KMT
women’s caucus and the cross party “Coalition for Promoting Women’s Rights.” The
KMT women caucus has been established since the first-term legislature. In the first
term of legislature, there was a coalition called the “League of Eight Ladies” which
included all the female legislators at that time (there were only 8 KMT women
legislators). It is a group that is looked upon as more of a social network, rather than as a
working team with any agenda. In the third-term. several KMT women legislators tell me
that they gather, on a semi-formal basis. almost once a month or every other month. in
one of the woman legislator’s district to sponsor her district’s activities. I asked most of
the KMT female deputies if any agenda were involved; they all told me that it was
activity-oriented. and thus did not possess any specific or serious agenda. The themes of
the activities could be various. and were decided by the local situations and the hostess
legislators. The main purpose of the meeting is to take turns rendering support to a
district activity hosted by one of the KMT women legislators. As one KMT female
legislator puts it.

For example. we went to the district A to participate in their musical or art activities

for the program of raising the standard of living in that district, or we went to district

B to participate in a talk with family women about parenthood. Its aim is to let their

voters see that. once you choose me. at the same time, you get some other people.
and we help each other out (interview: 98002).

Such a caucus is important for establishment women to socialize and receive support
from one another. Still other caucuses are important in as far as they enable women to
work together on legislative issues of mutual concern to them. Another women's caucus
within the Legislative Yuan was established in 1996. As previously mentioned. after the
murder of Peng, Wan-Ru. six DPP female legislators initiated the “Coalition for
promoting women’s rights”, which was endorsed by KMT and NP women legislators as
well. Seventeen out of the 21 women legislators all joined the coalition (all DPP female

members and all NP female members, excepting one, signed at the initiation). They

172

demanded the establishment of women’s rights committees in local and central
govemments. an increase in the budget and level of personnel dedicated to women’s
welfare. and establishing of a committee for gender equality in education and a
committee for preventing sexual assault. The urgent issues as regards this coalition, then.
concerned women’s safety.

Given that the coalition was a cross-party endeavor. in order to avoid any political
dispute. one of DPP initiators of the coalition pointed out that there is no chairperson. nor
any other formal position. Rather. the women simply gather together to campaign for
women’s issues. And in the beginning. they accomplished many things, as one DPP

initiator of the coalition recalls:

When it was just formed. we visited the Department of Health [to discuss women’s
issues]. They had never seen as many ten women legislators come together to talk
about these issues. and so then they began to consider them serious. If ten
legislators come together to a department. they will definitely feel enormous
pressure. It is so powerful that they do not dare to block it. but must work with us
seriously (interview: 98039).

Since the establishment of the Coalition (1996). many women’s bills withheld in
previous legislative terms have been passed in the third term. However. this initiator
(interview: 98039) also admits some problems are present within the coalition. Now that
the coalition has been operational for a good while. not everyone is as enthusiastic about
it. and there have been differences expressed upon various issues and problems. Only a
few legislators understand the problems completely and are acquainted with the goals and
direction of the coalition. the majority simply attends in order to endorse the cause. The

underlying reason for these differing degrees of involvement with the coalition is party.

and she adds:

There was certain period of time, during which some people were simply not
interested [in the coalition]. For example, if the KMT’s female legislators felt that
the coalition was led by the DPP. they would become less likely to join it. If that’s
the case. it does not matter. We do not have to use the Coalition to do our business.
Especially. right before the election, the strong party divisions would appear.

173

Anyway. the coalition is a form ofoperation and, whenever important issues arise.
we need to use the [leverage] power of the form. For example, to get the Law of
sex-invasion and the Law of road security passed, that’s what matters. Whether the
coalition exists or not. it does not matter. The next time we feel we need the clout of
the Coalition. we will operate again. For example. last year, representative Pang
advocated sex-invasion. but. in the beginning. she did not have the confidence that it
might pass. She did not gain support from her own party, and the administrative
departments were very reserved about that. It turned out that none of legislators
from three parties opposed it. because we forcefully negotiated to schedule it into
conference agenda and we passed it (interview: 98039).

On the men’s side. most do not have such a connection with women’s groups. Most
of them lack any sense of familiarity while mentioning women’s groups. at most they will
receive letters or petitions from them. A direct and close association has been rare. As
one male respondent (interview: 98026) points out, the women’s organizations have their
own set of myths. through which they confine their outside contacts to a few female
legislators. and he regards this as unwise.

In short, the majority of female legislators have established ties with women’s
groups, whether outside or inside of the Legislative Yuan, whereas most men have not.
This situation could possibly explain the gender differences that exist as regards
identifying with the female constituency and advocating women’s concerns. However.
the extent to which even women legislators are actually and actively engaged with these
women’s groups is highly varied. The predominant women’s network inside the
Legislative Yuan is the “Coalition for promoting women’s rights”, which does succeed in
connecting the majority of female legislators with issues of women’s rights, and does
work endlessly for women's interests. Nevertheless. inter-party divisions have prevented
it from becoming a strong and sustainable group, despite the fact that the coalition has
been the closest mechanism to an orchestrated establishment of women behind a common
cause so far. The other groups. which are active in lawmaking on women’s rights issues.
are the outside feminist groups. yet the relationship between them and congressional
women is remittent. Only a few female legislators have established cooperative and

steady relationships with outside feminist groups. The most stable and long-lasting

174

women’s network could well be found amongst the KMT's female legislators, both
outside and inside the legislature. Outside the legislative Yuan. the long history of the
connection between KMT women legislators and traditional women’s groups can be
ascribed to the KMT’s manipulation of local politics, through which they have
successfully co-opted most local traditional women's groups into the party. Inside the
Legislative Yuan. because most KMT women legislators are dispersed across different
electoral districts. competition among them is rare. With the women's league established
since the first term. the KMT women’s legislators have been semi-formally operating a
women's network. But the KMT’s women’s networks both inside and outside of the
Legislative Yuan can be deemed both conservative and traditional, and exist without a
common agenda. This also affects the type of female constituency KMT women
representatives identify with. and the issues with which they are concerned.

Finally. I try to summarize the stability of the association between women legislators
and women’s organizations, and whether the network involves the law-making function

in table 5-6.

Table 5-6. Women’s groups inside and outside of government

 

 

 

 

 

 

 

 

 

Stability Law-making
Cross-party women caucus No Yes
Within-party women caucus Yes (KMT) No
Feminist group No Yes
Traditional women’s group Yes No

 

Given the different types and degrees of association between women’s organizations
and congresswomen, it is difficult for me to prove directly or even to make a causal
relational generalization as to whether there is involvement with women’s culture and an

accompanying identification with a female constituency and with certain women-oriented

175

policy priorities. Ilowever. I have shown that women are. on average, more likely to be
involved with or develop women’s networks, and that such networking is conducive to
developing a stronger female group identity and identifying with women’s culture in a
variety of ways. As Carroll (1992) noted. women in male-dominated institutions need
access to a separate and alternative women’s community in order to sustain their
participation and to enable them to push for policies for women. From my interviews
with the third-term legislators. it seems fair to conclude that women in public office. who
are involved with women’s organizations outside or inside government, will be more

likely than men to represent female culture and to advocate the rights of women.

176

CHAPTER 6
CONCLUSION

This aim of this project is to examine gender representation in Taiwanese politics. I
began with a theoretical examination of the nature of female representation in Chapter 1.
and then proceeded. in Chapter 2, to depict the changing sociopolitical context of Taiwan.
setting the stage for the investigation of gender differences in Taiwanese representational
politics that followed. In the theoretical part, the dispute between liberalism and the
politics of difference concerns whether identity characteristics (such as gender) are
correlated with relevant political variables to be represented and whether a link exists
between descriptive representation and active representation. Liberalism emphasizes
individual equality and procedural faimess regardless of any notion of identity difference.
whereas proponents of the politics of difference argue that any real representation must
start with some recognition of the differences between identity groups. To resolve this
disagreement over the impact of identity on representational politics, I chose to test the
impact of gender identity on the legislative process. In addition. by organizing possible
gender differences into five hypotheses, further based on a behavioral and attitudinal
distinction, I have managed to carry out two separate examinations into gender
representation in the Taiwanese Legislative Yuan - namely. longitudinal behavioral
research covering 1990-1997, and attitudinal interviews with third term legislators.

Based on the hypotheses I put forward earlier, my conclusions are as follows.

177

Research findings

Htpotlzesis / : institutional committee assignment

In terms ofinstitutional arrangements. studies in the 19808 found that women were
more likely to serve on both the education and interior committees. However. in my
comparison. such a clear gender pattern of committee assignment has not been found in
the 19905. Though women are still more likely to join the education committee. the
degree of difference between the two genders as regards other committees is not
significant. For example. the interior committee has turned out to be one of the most
popular committees since. following democratization. it has concerned itself with a wide
range politically important bills. Therefore, both men and women are now equally
interested in joining it. Moreover. women also actively participating in the defense
committee - a committee traditionally considered something of a "men’s club."

The reasons for the blurring of gender division as regards committee choice are
many. Firstly. a variety of factors affect legislators’ committee choices, such as the
interests and urgings of their party. and the individual’s own self-interest and
constituency considerations. Clearly. a final choice of committee is the product of each
legislator's considered weighing of a number of factors. and gender is, for many. simply
one of such consideration, and often not the primary one. Moreover, given that the social
and professional opportunities for women have expanded over time to include non-
traditional occupations. it should perhaps not be surprising that recently elected women
would select a wider range of committee assignments. Secondly. the committee system is
not well institutionalized in Taiwan and certain "rules" that exist in other countries, such
as those regarding seniority and specialization, have not yet established themselves in the
Legislative Yuan. Consequently, legislators tend to move amongst different committees

in order to gain influence in various fields. and only a few legislators would claim any

178

form of authority over certain committee areas. Thirdly, because of jurisdictional
overlap. an examination of the committee system is not in itself sufficient to reveal
legislators’ real policy interests. Committee participation can only, at most, serve as a
general indicator of the policy interests of the legislators. Given this institutional
flexibility and the constraints of the party system, the choice of committee appears to

show a fluid and random pattern in Taiwan. largely unrelated to gender.

Hypothesis 2: legislative participation

 

Policy responsiveness involves a prolonged process of legislative activities, many of
which cannot be determined by reference to the committee system. A legislator can be a
committee member without really paying much attention or giving a great deal of time to
the bills examined in the committee. To counter this. I reclassified legislative activities
according to the subjects of the bills and then examined the gender patterns of legislative
participation in different policy subjects at the different stages of legislative participation:
i.e.. initiation. chairship. speech at the committee stage, and speech at the final Yuan
stage.

By treating women and men as two different groups. I compared the policy priorities
of each at different stages of the legislative process, on both an aggregate and individual
level. On the aggregate level, female legislators appeared to participate most often in the
caring and communication policy area, whereas men legislators appeared to participate
most often in the economic or development policy area. This pattern of difference was
found across different stages - from agenda setting to policy deliberation - and across a
number of legislative terms. The only exception was in the second term. when female
legislators were most likely to initiate bills in the area of economic development.

I also compared the gender gap in terms of the percentage of legislator participation
by each group in caring and communication and feminist issues. by deducting the

percentage of male legislator participation in these policy areas from that of female

179

legislators. In this regard. too. the findings are consistent with my hypothesis: namely,
that women tend to be more involved in these particular policy areas than men do.
Furthermore. at the individual level. I have shown that gender is significantly and
positively associated with the degree of legislative participation in these two policy areas.
The OLS regressions. after controlling other factors that might affect the degree of
legislative participation (e.g.. age. education. party, constituency). also suggested that
gender is both a positive and, in general. significant factor in accounting for the variances
between the sexes in these policy domains. This positive association between gender and

participation in these policy areas is especially clear in the third term.

Hypothesis 3 .‘ constituency perception

 

In approaching legislators’ own subjective perception of whom and what they
represent. the testing of hypotheses three and four required the conducting of a series of
interviews with a matched sample of 40 third term legislators. By separating their
constituencies into several groups, I asked legislators who they considered their primary
constituency to be. Examining the results of the interviews. I discerned several patterns.

Firstly. legislators' identifications of their own constituency are influenced by a
variety of factors. These include such issues as electoral rules, constituency demography.
seniority in office. previous profession, class, or gender. and combine together to shape
an individual legislator's evaluation of the relative importance to them of different
constituencies. Given the overlapping characteristics or identity that each legislator
might possess. a somewhat mixed. as opposed to entirely cohesive, concept of
constituency came across in the interviews. Evidently. gender cannot easily transcend
matters ofclass or other identity in defining whom a legislator will see as their primary
constituency.

Secondly. when the respondents were divided into two gender groups. a significant

gender difference became apparent as regards the degree of identification present with

180

three particular groups: namely. children. the elderly, and women. Female respondents
were more likely to identify with these groups than men were, and the difference was
especially striking as regards the level of identification with "women" as a group. Over
halfof the female legislators identified with their female constituency, whereas no men
did so. This is consistent with the theoretical expectation that the female's primary role in
the family as a caregiver has in some ways been transferred into the representational
arena.

Thirdly, men tended to be less group-oriented than women. Among all the
categories of group constituency. only as regards the categories of the middle-class.
business people. and environmental groups, did men outnumber women in reporting these
categories as their constituency groups. Also. more men than women respondents denied
the general idea of group constituency, considering the idea of group interest as partial
and subjective. Instead they insisted that the majority interest, being unbiased and
neutral. should always be the primary concern of the legislator. As I have already pointed
out. such a viewpoint reflects a liberal approach to representation.

Finally. from the different self-reported constituency identifications. I further
explored legislators’ constructive willingness to represent the “women’s interest” so as to
examine whether group identification is linked to an active and actual commitment to that
group. Based on respondents’ observations and experiences, there was a near across-the-
board consensus that women were more actively involved in the representation of

women’s issues than were men.

 

Hts/pothesis 4: policy priority

At least two ways to examine legislators’ degree of commitment in the
representation of certain groups are available to us: firstly, we can observe their actual
legislative participation in different policy areas over the whole legislative process or.

alternatively. we can cross-examine the legislators themselves. The former approach was

181

employed in the testing of hypothesis two and a general pattern of gender difference was
discovered. The latter approach was used in hypothesis four to provide a more specific
reference to the actual setting of agendas, which is the first step of the legislative process
and requires the most commitment and expertise.

Generally. the answers I received delivered the following messages with respect to
the agenda setting. First of all. there is indeed a gender gap as regards the identification
of legislative priorities. Women’s issues topped the female list of policy priorities.
whereas no man described this area as being amongst their policy priorities. However.
the differences in other policy areas were not significant. For example, more women than
men listed caring and communication issues as policy priorities. but the discrepancy was
not significant. And regarding development issues, the frequency of this being reported
by female legislators as at the top of their agenda was even greater than that of male
legislators. In the policy areas of defense and foreign affairs and politics and
government. this arrangement was reversed. Overall, women appeared to report social
and domestic issues as their top policy agenda. whereas men were more likely to attend to
national and controversial political issues.

Secondly, I also asked whether the change in the gender composition of the
Legislative Yuan would bring agenda change. Consistent with the above findings. most
respondents who actually agreed that there would be a change of agenda following the
introduction of more women. expected that change to lead to an increased emphasis upon
women's interests. However, most were not convinced that there would be any such
substantial alteration in the existing legislative agenda at all. Female legislators
perceived at least three practical and psychological constraints to this: namely, the ways
in which female legislators differ amongst themselves, the fear of being labeled as
narrow-minded or blinkered, and the passivity of female voters as regards supporting the
actions and proposals of female legislators. These obstacles account for the difficulty one

faces in linking the mere descriptive presence of women to the substantive representation

182

of women’s interests.

Hypothesis 5 .' the consistency of gender difference across different legislative

 

dimensions

In hypotheses one to four. I have examined gender differences apparent in various
legislative dimensions: institutional committee arrangements. actual legislative
participation. constituency perception. and subjective policy priorities. With the
exception of the first of these dimensions. the existence ofa definite gender pattern has
been discerned throughout. To discover whether these gender patterns are random, or
whether they represent something more significant. I tested the association between
objective legislative participation (hypothesis two) and the subjective identification of
policy priorities (hypothesis four). Then I looked for a correlation between constituency
perception (hypothesis three) and these declared policy priorities (hypothesis four).

First of all. I examined the possible link between the objective measure of legislative
participation and the subjective perception of legislative priorities by examining the
correlation of the third term legislators’ performances on these two dimensions. Among
the seven policy areas that I classified, and with only a few exceptions, the pattern of
actual legislative agenda setting (i.e.. bill-initiation and chairship) is positively and
significantly associated with that of declared policy priorities. This demonstrates a
consistent relationship between legislators’ declared policy priorities and their actual
legislative behaviors.

Secondly. I examined the association between legislators’ perceptions of their own
constituency and. once again, their declared policy priorities. In each dimension, a
gender pattern was detected in which women were more likely to identify with a female
constituency and also listed women’s issues among their "top-three" policy priorities.
Assuming that the constituency grouping with which legislators are most concerned

might have an impact on the policy issues upon which they will spend time. I examined

183

this causal postulation via the use ofa logistical model. After controlling other factors
that might influence a legislator's choice of policy priority. the results demonstrated that
whether legislators identified themselves with a female constituency or not, significantly
affected the chances of "women’s issues" being a primary policy agenda for that
legislator.

Ultimately. in order to explain the consistent gender pattern that I have
demonstrated. I utilized Carroll’s (1992) "cultural" explanation. The growing influence
of women’s associations, and the increasingly close relationships between female
deputies and these various women’s organizations. have played a major role in bringing
about the female legislators’ general affiliation with women as a group and their deeper
commitment to representing female culture. By examining the diversity and dynamics of
the female legislators’ association with women’s organizations, the differences that exist

among female legislators themselves could be further recognized

The implication and discussion

Compared with past studies of female compatibility with men in the arena of
representational politics, this study goes further. and looks to explore exactly what it is
that distinguishes women from men, and the rationale of considering gender divisions as
an integral part of normal party politics. If women represent essentially the same
constituency or the same policies in representative politics as men. it seems redundant or
simply symbolic to describe women as a "category" in that arena. Representative politics
differs from merit-based and expertise-oriented administrative systems in so far as it
means to represent the people. Therefore. it is reasonable to inquire exactly who
represents whom and what.

I began by sketching the institutional and sociopolitical background of Taiwan.
which has provided a favorable environment for an ever-increasing level of attention

being devoted to women’s issues. Building upon this background. my research has

184

demonstrated that identity characteristics are indeed correlated with representative
politics. and that legislators’ gender identity affects their legislative behaviors and their
subjective perceptions. The behavioral research regarding legislative participation from
agenda setting (bill-initiation and chairship) to policy deliberation (speeches in committee
and on the floor) has shown the different policy interests of male and female legislators
across multiple terms. Women have consistently placed more emphasis on the policy
areas of caring and communications and women’s issues than have men, and men have
tended to place more emphasis on political and economic issues. The behavioral research
based on the objective measurements was then attested from the input side - i.e..
legislators’ subjective attitudes. The attitudinal research that followed - the interviewing
of legislators so as to elicit their constituency identification and policy priorities - has
further confirmed this gender pattern. It has especially pointed out the significant gender
differences in the agenda setting process from legislators’ own observations. Finally,
these findings have. as a whole. paved the way to establishing a link between subjective
perception and objective legislative participation. along the lines of gender. At the same
time. from the interviews. the psychological obstacles that can prevent legislators acting
upon their own differing agendas were also documented. These obstacles determine the
extent to which subjective identity perceptions will lead to different and gendered levels
of legislative participation.

The group identity and common agendas of female legislators seem to embrace and
accommodate both their traditional roles in the private domain and a feminist concern for
women’s rights. The identity characteristics generated from female psychological and
sociological experiences have been transferred into representational politics. The linkage
between descriptive representation and substantive representation discerned in this study
buttresses the assertion that who is present in legislative assemblies is closely related to
what legislators actually do. Accordingly. it would appear that the “politics of idea”

cannot be separated from the “politics of presence” along gender lines. In other words.

185

this research has demonstrated that gender characteristics are correlated with
representative politics. This correlation carries both practical and theoretical
implications.

In practice. the analysis of the legislative behavior of female members and their
responsiveness to women could support the justification for increasing the number of
female representatives. As the number of women elected to the Legislative Yuan
continues to increase. we might expect this increase in the substantive representation of
women’s interests to continue and the representative gap between women and men to
continue to escalate. While discussing gender quotas in Taiwan, one male legislator
claimed that the reason he supported increasing gender quotas was because, generally.
men are less concerned than women about women’s issues, and so the system requires
more women to enter the process to redress this (interview: 98024). Female
representation. rather than simply visibly representing ideas of gender equality, implies
substantive change in legislative processes and agendas, even if the degree of change is,
as discussed above, still subject to many other considerations.

Furthermore. the findings also shed light on the theoretical dispute between the
theory of liberalism and the politics of difference. with the former emphasizing the
unimportance of identity difference. and the latter claiming that difference counts. My
research cannot sustain either of these contentions in their entirety, which was also not
my intent. Nevertheless. it does support the empirical basis of the politics of difference
as regards representative politics. Gender division is a significant variable in explaining
the specific policy choices and the general nature of representative politics. In advocating
the equality of individual representation with no regard for the significance of group
difference and representation. liberalism overlooks the implications of group identity to
representative politics. Identity characteristics could be a necessary, though not
sufficient. factor in bringing about a structural change in the social agenda of

representative politics. as would appear to be the case in certain Nordic countries. where

186

increasing female representation has introduced more social welfare concerns into the
political agenda. In short, this would suggest that the linkage between identity and

interest should be properly addressed in future studies of legislative and electoral politics.

Suggestions for future works

Above all. several points remain which must be handled with caution in future
research. Firstly. any research should try to avoid the danger of "essentializing" gender
differences without first examining the sociopolitical and institutional contexts in which
they exist. In discussing gender differences in politics, or in any social sphere, a tendency
to overlook the differences that are derived from other social separations is always
present. Indeed, one must first admit that there are no generic women versus generic men
situations. rather, it is a matter of specific women versus specific men in different
institutional and sociopolitical contexts. Failure to take into account the interaction
between gender and other sociopolitical divisions will inevitably threaten the possibility
of understanding gender dynamics in their sociopolitical contexts. My study has
attempted to take the party factor into account by controlling its influence. however. a
more dynamic and in-depth exploration of the interaction between gender and other
sociopolitical factors could be even more inspiring.

In addition, many of the legislative procedures have been experiencing change.
Since the start of the fourth term in 1999, the legislative process has become increasingly
complex, because of the increasing number of legislative members. the increasing
number of factional members in the legislative Yuan. the changing party politics. and
actual alterations to legislative procedures. In terms of this last, a change to the party
proportions regarding committee assignment, and the formal institutionalization of party
negotiations concerning bills will enhance party discipline over individual members and
thus the party's dominant role in the legislative process. More attention must be paid to

the new legislative dynamics caused by different compositions. procedures. and party

politics.

Secondly, because my primary interest was sorting out the different policy interests
of representatives. 1 compared gender differences by computing the intensity of
behavioral participation in different policy areas. The behavioral measurements I used
cannot reveal the possible policy position of legislators’ interest within each policy
domain. The whole legislative process from agenda setting to policy formation cannot be
fully understood without knowing each legislator's viewpoint regarding each issue. For
example. two legislators can participate actively to the same degree on the same issues
but in a totally opposite direction. My study can only reach conclusions regarding the
varied degrees of legislative participation and attention devoted to different policies. not
the concrete policy positions of the legislators. Further exploration of policy direction
and content would certainly improve this study.

Thirdly. this study did not spend much time discussing another important dimension
of representation - constituency service. In the literature review. I have covered several
U. S. studies and one recent Taiwan study upon legislators' constituency service (Huang
1994). Though their findings - that sex is a significant factor in accounting for the
discrepancy of levels of constituency service undertaken between women and men - is
consistent with my theoretical projections. I have criticized the methods and the
representativeness of the samples used. Because most constituency service is passed
down to legislators’ local agents. the numbers obtained from a survey of these legislators’
assistants (Huang 1994) cannot determine the direct connection between legislators and
their constituency. In my interviews, I attempted to ask legislators directly about their
efforts in constituency service. however, many of them failed to answer. This certainly
does not suggest that we should give up the attempt to understand legislators’
constituency service patterns. but, rather. that we should discover alternative methods
(e.g.. participant observation as F enno did or field investigation) with which to do it.

Another complete and well-designed study will be required for this.

188

Furthermore. this study focuses primarily upon the supply side of the legislators’
methods and degrees of policy representation without paying sufficient attention to the
demand side of the equation - i.e.. to the voters themselves. Though I have mentioned
some of the voters’ policy interests studied in the past (Fan and Hsu 1994), they did not
fit with the policy classifications I have used in this legislative study. Most studies of the
level and form of policy representation by female legislators regarding their female
constituencies have been based on theoretical assumptions of women’s sociopolitical
experiences and past studies. Also most of the legislative study on gender difference has
failed to examine and incorporate the policy demands that emanate from the constituency
and the electoral base of the representatives. Overall then. more accurate data from the
input side of policy formation. such as concerning the constituency's own preferences and
degree of electoral support. would be constructive in tracing the correspondence between

a female constituency and female representatives.

189

APPENDIX 1

190

APPENDIX 1

Bill classification

First stage:

I commenced by dividing the bills into 19 coding schemes. For each bill examined

in the committee meeting. one or two subjects and types are assigned for each meeting in

the Yuan index record. This was my most important source in setting up the first-stage

scheme. 1 present all the categories and some examples of the subjects of bills that fall

within those respective categories as:

1.

2.

10.
ll.

12.

Economic Affairs and Finance: taxes. finance; stock market, energy, licensing.
Budget and Local development: budget. transportation (highway, airport. port, post).
public construction. local development. tourism development. land development.
Political issues: elections. political reforms. institution reform. ethnicity issues.
mainland china policy.

Government Organization: government personnel. recruitment. merit system of
public servants. infrastructure.

Education and Culture: school administration, teacher credentials and training.
museums. management of cultural heritage.

Mass-communication: TV, PBS, radio. newspaper. magazines.

Judicial system: prosecution, lawyer, civil law, criminal law, law enforcement.
Social order: gun control. public building management, policemen. personnel.
smuggling, game store management, stall-keeper management. drugs.
Environmental: air pollution. water pollution.

Medical : health system. public sanitary. hospital management.

Labor: wages, unions. hiring and hiring. labor health care. labor training programs.
labor pension.

Agriculture: farm. food and livestock. senior farmers’ pensions.

191

l3. Fishery: ship management.

14. Women’s right issues: gender equality at work. sexual harassment. sexual invasion.
women’s rights in civil laws. marriage. property rights.

15. Children: child prostitution, child care. youth issues.

16. Ethnic minority: aboriginal land issues. language education, names.

17. Welfare: health care. enforced comprehensive car insurance.

18. Foreign Affairs: embassy appointments. international relations. international
economic cooperation.

19. Defense: military school. military court and prosecution, veterans.

Second stage

Next. I further collapse the bills into a more macro-structured scheme. Three policy
schemes are considered as references here. F irstly. because of my interest in exploring
the gender divisions that affect legislative participation as a whole, I refer to Thomas and
Saint-Germain’s gender schemes for collapsing policy. According to their classification.
the categories of education. children, health. and welfare are traditionally regarded as
being "women’s interest" categories. I collapse these categories with the domain of
communication and then label it the care and communication domain. Another issue area
that I theoretically hypothesized to be gender-specific is the women’s rights issue. As to
the rest of the categories. I tried to put those categories with most similar characteristics
together by referring to two general policy schemes.

The next scheme I referred to was Sheng’s scheme (1996) for studying the factors
influencing legislative participation in Taiwan. Sheng divided the bills into political
affairs. economy and finance. government organization, social welfare. education and
communication, judicial matters and social order, and environment protection. She
. classified bills by the content of the bills under examination, not by the name of the bills.

Since she only sampled some bills from those that were proposed and passed on the

192

second and third term. her scheme was sufficient to cover and analyze those bills. If one
were to attempt a full coverage of all the bills discussed, however, then it would soon
become clear that some bills did not fit into any of Sheng’s categories. For example.
foreign affairs. defense, minority. and medical bills were not included in her scheme.

The last reference I have extensively used in this regard is Chou, Clark and Clark’s
classification of resolutions in studying women’s representatives in Taiwan. They
classified bills into economic affairs. public works. local development, education and
culture, elections and civil and human rights. foreign affairs, and defense. The problem,
as above, arises from the exhaustibility of their scheme. And the way Chou, Clark and
Clark determined the type of bills was by asking legislators to list all the bills they
sponsored in each category researchers had pre-set. In this study. I want to know
legislators’ subjective policy interests among all of the legislation. It would be easy to
induce the classifications from them. yet. it is impossible to have all legislators attempt to
classify all the bills they attend to.

According to these references, I collapsed the original 19 policy categories into the

seven categories as shown in table appendix 1

Table appendix 1. Two stages of bill classification.

 

 

 

 

 

 

 

 

 

 

First-stage 19 categories Second-stage 7 categories
Education. Children. Health. Welfare. Communication Caring and Communication
Women’s rights issue Feminist issues

Finance. Budget. Economics Economics and Development
Judicial. Social Order Judicial and social order
Defense, Foreign Affair Defense and Foreign affairs
Politics. Government Organization Political and Government
Labor, farmer, fishermen. ethnicity minority. Group interests

environment group;

 

 

193

APPENDIX 2

194

APPENDIX 2a

Interview cover letter

Dear Representative

This project attempts to gain insights into the workings of legislature through
interviews with its members. It is a part ofa dissertation project on Taiwanese
representative politics by Wan-Ying Yang. a Ph.D. candidate at Michigan State
University. Through her training at MSU for the past four years. she has developed a
research interest in democratic processes and institutions. as well as representative
politics across countries. including Taiwan. As members of her dissertation committee.
we humbly request that you consider joining this project by participating in an interview.

As we know. Taiwanese democracy has made much progress since 1986. For a
consolidated democratic country. a well functioning legislature plays an unsubstituted
role. It is. therefore. in our highest interests as political scholars. to learn from people
who practice and participate the representative system on a daily basis. Your experiences
in the legislature will surely provide us with valuable information about how Taiwan’s
representative system functions.

We do hope that, through this project and your kindly assistance, people who are
concerned about democratic politics in Taiwan. may be able to have a more in-depth and
contextualized understanding about the operation of representative politics in Taiwan.

We thank you in advance for your consideration.

Sincerely.
Dr. Jones. Mark
Patterson, Dennis
Rohde. David
Department of Political Science.

Michigan State University.

195

 

APPENDIX 2b

Interview questions guide

The purpose ofthis interview is to gain insight into the workings oflegislators. We

assure complete confidentiality of answers.

General perception on legislative activities.
The first set of questions deals with your views about performing some aspects of your

jobs.

In your view, what is your main duty as a legislator?

( a suggesting list of legislative works. for example: making good policy. holding public
hearing. serving constituency. working with interest groups)

Approximately how many hours a week. on average. do you devote to your job as a
legislator working in the process of making policy (including policy debate, initiation, co-
sponsoring. voting for policy. etc.) and serving your constituency (including casework,

attending local community activities. etc)?

Of this amount. how much. as a percentage. of your time do you spend on making policy

(including policy debate. initiate. cosponsor. vote for policy. etc)?

Of this amount. how much. as a percentage. of your time do you spend on serving your
constituency (including looking after constituency, casework; attending local community
activities. etc.). approximately?

What are the important factors that affect the success of getting bills passed?

Is there any position in the Yuan that is important to gain influence?

Perception of representation of constituency (Hypothesis 3).

As a representative, you represent many different constituency groups. Do you consider

yourself representative of certain constituency groups (certain kinds of people, organized

196

or unorganized. within or outside your own district)?

----- ( If no. probe.) Do you consider certain group’s concern as very important and
worthy of being addressed by you? For example, people in your district, a certain ethnic
constituency, a single-issue group( e.g., environmental). a business constituency. labor
groups. women, elderly. children. poor. or others.

----- (if yes) who are they?

Did you gain support from these groups in your last campaign? 18 there any groups that

constitute your strongest source of support? Second strongest source of support?

Do you find it easier to represent these groups’(people’s) concerns? Why/why not?

Do you feel you are better able than most other legislators to represent these groups’

(people’s) concerns?

----- (if no) Are there any groups you find easier to represent than others?

Perception of policy priorities (Hypothesis 4)

This set of’question deals with your legislative priorities:

Please list your top three priority bills in the last complete legislative session?

How do you choose your legislative priorities?

Why are these issues important to you?

Do you make contact with outside groups while advocating these issues?

Whom do you ask support from for these issues?

What are the difficulties for you to promote these issues in the Yuan?

Do you spend most of your time on them? Or do you distribute your time evenly among

197

most policies?

What are the accomplishments you are proudest ofduring your term as a legislator?

Women in politics (related to hypothesis 3, but put in the last part to avoid leading

respondent.
Recently, the discussion on women quotas in electoral rule is still under debate.
[)0 you think that having more women in representative politics would make a difference

to the legislative process?

Do you think that having more women in the representative politics has made a difference

in the legislative process?

---- If yes. what is the most important way women have affected the politics?

---- In what other ways. have women made a difference?

---- Are you particularly interested in seeing more women in politics? why or why not?

Do you feel that women as a group have particular concerns which the government

should address or is addressing ?

---- If no. why not? ( finish this part)

---- If yes. what are some of those concerns?

(For Female) Do you think you and your female colleagues are better able than your male

colleagues to represent such concerns? If yes, why?

(For Male) Do you think your female colleagues are better able than your and your male

colleagues to represent such concerns? If yes, why?

(For Female) Do you think you and your female colleagues are more willing than your

male colleagues to represent such concerns?

198

(For Male) Do you think your female colleagues are more willing than you and your male

colleagues to represent such concerns?

Do you think there are any barriers specific to women in elective politics? ( Probe: are

there “rules of the game” that particularly apply to women? )

199

BIBLIOGRAPHY

200

BIBLIOGRAPHY

Amy. Douglas J. 1993. Real Choices/New Voices: the Case for Proportional
Representation Elections in the United States. New York: Columbia University
Press.

Antolini. Denise. 1984. Women in Local Government: An Overview". In Political

WomenxCurrent Roles in State and Local Government. edited by J. F lammang.
Beverly Hills, CA: Sage.

Baxer. S. and M Lansing. 1983. III/"omen and Politics. Ann Arbor: University of Michigan
Press.

Berg. John C. 1994. Unequal Struggle: Class, Gender, Race, and Power in the U. S
Congress. Boudler: Westview Press.

Boles. Janet K. 1991. Advancing the Women's Agenda Within Local Legislatures: The
Role of Female Elected Officials. In Gender and Policymaking: Studies of Women in
Office. edited by D. L. Dodson. New Brunswick. N.J.: Center for the American
Woman and Politics.

Bosco. Joseph. 1994. Taiwan Factions: Guanxi, Patronage. and the State in Local
Politics. In The Other Taiwan: 1945 to the Present. edited by M. A. Rubinstein.
Armonk. N.Y.: ME. Sharpe.

Brodie. J ., and L. Gotell. 1991. Women and Parties: More Than an Issues of Numbers. In
Party Politics in Canada, edited by H. G. Thornbum. Scarborough, ONtario:
Prentice-Hall.

Bryson. Valerie. 1992. Feminist Political Theory. New York: Paragon House.

Burrell, Barbara C. 1996. A Woman 's Place is in the HousesCampaignfor Congress in
the Feminist Era. Ann Arbor: University of Michigan Press.

Bystydzienski. Jill M. 1995. Women in Electoral Politics: Lessonsflom Norway.
Westport: Praeger.

Bystydzienski. J. M.. and Y. Lin. 1994. Women in Politics in Norway and Japan: A
Comparative Study of Peace Issues. In Peace and Japan, edited by S. Takayanagi
and K. Kodama. Kamihama: Mie Acadamic Press.

Carroll, Susan. 1984. Women Candidates and Support for Feminist Concerns: the Closet
Feminist Syndrome. Western Political Quarterly 37:307-23.

Carroll. Susan, J.. D.L. Dodson. and RB. Mandel. 1991. The Impact of Women in Public
Office: CAWP.

C arroll, Susan J. 1985. Women as Candidates in American Politics. Bloomington:
Indiana University Press.

Carroll. Susan J. 1992. Women State Legislators. Women's Organizations. and the
Representation of Women's Culture in the United States. In Women Transforming
Politics. edited by J. M. Bystydzienski. Bloomington and Indianapolis: Indiana
University Press.

C hao. Hsiang-Chun. 1995. The Study of Taiwan's Women's Voting Behaviors. M.A.
Thesis ( in Chinese). Soochow University. Taipei, Taiwan.

Chen. Ming-Tong. 1994. Looking for Faction Voters. Paper read at Democratization.
Party Politics. and Election.

C hen. Ming-Tong. 1996. Local Factions and Elections in Taiwan's Democratization. In
Taiwan 's Electoral Politics and Democratic Transition." Riding the Third Wave,
edited by H.-m. Tien. Armonk, N.Y.: ME. Sharpe.

C heng. Chi-hui. 1983. What Have Legislators Done for Women? Women’s Magazine. 24-
30 ( in Chiness).

C heng. Li-ping. 1997. A Policy Analysis on Gender Equality Related Bills by the
Second- and Third-term Legislators. M.A Thesis. Soochow University ( in
Chinese). Taipei, Taiwan.

C hiang. Su-hui. 1996. The Study of Female Elites--the Cases of Democratic Progressive
Party's Female Officials. M.A. Thesis ( in Chinese), National Taiwan University,
Taipei. Taiwan.

Chou, Bih-er. 1994. Changing Patterns of Women's Employment in Taiwan, 1966-1986.
In The Other Taiwan: 1945 to the Present. edited by M. A. Rubinstein. Armonk,
N.Y.: ME. Sharpe.

Chou. Bih-er. 1995. The Hen Crowing at Dawn ( a Woman Usurping Man's Power)--
Women and Politics. Paper read at Women's Study in Ten Years-«Retrospect and
Prospect of Women's Rights, at Taiwan University ( in Chinese).

Chou, Bih-er, Cal Clark, and Janet Clark. 1990. Women in Taiwan Politics: Overcoming
Barriers to Women 's Participation in a Modernizing Society: Lynne Rienner
Publishers, Inc.

C hu. Yun-han. 1994. Social Protests and Political Democratization in Taiwan. In The
Other Taiwan31945 to the Present. edited by M. A. Rubinstein. Armonk, N.Y.: M.
E. Sharpe.

Chu. Yun-han. and Tse-men Lin. 1996. The Process of Democratic Consolidation in
Taiwan: Social Cleavage. Electoral Competition, and the Emerging Party System. In
Taiwan 's Electoral Politics and Democratic Transition.“ Riding the Third Wave,
edited by H.-m. Tien. Armonk, N.Y.: ME. Sharpe.

Clark. Janet. and Cal Clark. 1997. Women and Democratization in Taiwan. Paper read at
XVII World Congress of the International Political Association, at Seoul.

Dahlerup. D. 1988. From a Small to a Large Minority. Scandinavian Political Studies
I 1:275-299.

D'Amico, F. 1995. Women National leaders. In Women in World Politics. edited by F.
D'Amico and P. Beckman: Bergin and Garvey.

Darcy. R.. Susan. Welch, and Janet. Clark. 1994. Women, Elections, and Representation:
University of Nebraska Press.

Davidson. Roger H., and Walter J. Oleszek. 1996. Congress and Its rifen-zbers. 5th ed.
Washington. DC: Congressional Quarterly Inc.

Diamond. I. 1977. Sex Roles in the State House. New Haven: Yale University Press.

Dickson. Bruce J. 1996. The Kuomintang before Democratization: Organization Change
and the Role of Elections. In Taiwan 's Electoral Politics and Democratic
Transition: Riding the Third Wave. edited by H.-m. Tien. Armonk, N.Y.: ME.
Sharpe.

Dodson, Debra L. 1991. Gender and Policymaking: Studies of Women in Office. New
Brunswick, NJ: Center for the American Woman and Politics.

Duerst-Lahti. Georgia, and Rita Mae Kelly. 1995. The Study of Gender Power and Its
Link to Governance and Leadership. In Gender Power, Leadership, and
Governance. edited by G. Duerst-Lahti and R. M. Kelly. Ann Arbor: University of
Michigan.

Engstrom, Richard. 1987. District Magnitudes and the Elections of the Irish Dail.
Electoral Studies 6: 123-132.

Epstein. C. F. 1983. Women in the Law. NY: Anchor.

Fan. Yun. and Yung-min Hsu. 1994. The Discoverred Women's Voters. Paper read at

203

Democratization. Party Politics. and Elections. at National Taiwan University.

F arris. Catherine S. 1994. The Social Discourse on Women's Role in Taiwan: A Textual
Analysis. In The Other T aiwan51945 to the Present, edited by R. M. A. Armonk,
N.Y.: ME. Sharpe.

Fenno, Richard F. 1978. Home Style .° House Members in Their Districts: harper Collins
Publishers.

Flammang. Jannet A. 1985. Female Officials in the Feminist CapitalzThe Case of Santa
Clara County. Western Political Quarterly 38 (94-1 18).

Frankovic. K. 1977. Sex and Voting in the US House of Representatives. 1961-75.
American Politics Quarterly 5:315-30.

Freeman. Patricia K.. and Lilliard E. Richardson. 1995. Gender Differences in
Constituency Service Among State Legislators. Political Research Quarterly 48

(1):169-179.
Galligan. Carol. 1982. In a Different Voice. Cambridge: Harvard University Press.

Gallin. Rita. 1984. Women. Family. and the Political Economy of Taiwan. Journal of
Peasant Studies 12 (1)176-92.

Gehlen, FL. 1977. Women Members of Congress: A Distinctive Role. In A Portrait of
Marginality, edited by M. Githens and J. Prestage. New York: McKay.

I-Iall. Richard. 1996. Participation in Congress. New Haven: Yale University Press.

Havens. Catherine M.. and Lynne M. Healy. I991. Cabinet-Level Appointees in
Connecticut: Women Making a Difference. In Gender and Policymaking: Studies of
Women in Office, edited by D. L. Dodson. New Brunswick. NJ: Center for the
American Woman and Politics.

Havvio-Manilla, Elina et al. 1985. Unfinished Democracy: Women in Nordic Politics.
New York: Pergamon.

I—Iickman. John C. 1997. The Candidacy and Election of Women in Japanese SNTV
Election Systems. Women and Politics 18 (2):1-25.

Ho. Szu-Yin Benjamin. 1986. Legislative Politics of the Republic ofChinazl970-l984.
Ph. D. Dissertation, University of California, Santa Barbara.

Hu. F o, and Der-Yu Chen. 1987. Voting Behaviors. Taipei: Central Electoral Committee
( in Chinese).

204

Huang. Shiou-Tuan. 1994. Constituency Service: The Basis ofthe Re-electionfOr The
Legislators. T aipei, Taiwan: Tung-Shan ( in Chinese).

IPU , Inter-Paliamentary Union . 1992. Women and Political Power: Survey Carried Out
among the 150 National Partliaments Existing as of 31 October 1991. Report and
Ducuments series 19.

Johnson. Janet Bottolph. and Richard A. Joslyn. 1991. Political Science Research
.ilvlethods'. Washington. DC: Congressional Quarterly Inc.

.lonasdottir, Anna G. 1988. On the Concept of Interest. Women's Interests, and the
Limitations of InterestTheory. In The Political Interests of GendersDeveloping
Theory and Research with a Feminist Face, edited by K. B. Jones and A. G.
Jonasdottir. London: Sage Publications.

Kanter. Rosabeth M. 1977. Some Effects of Proportion on Group Life:Skewed Sex
Ratios and Response to Token Women. American Journal o/‘Sociology. 82:965-90.

Karvonen, Lauri. and Per Selle, eds. 1995. Women in Nordic Politics: Closing the Gap:
Dartmouth Publishing Company.

Kathlene, Lyn. 1989. Uncovering the Political Impacts of Gender: An Exploratory Study.
Western Political Qarterly 42:397-421.

Kelly. R.M., Saint-Germain M.A., and JD. Horn. 1991. Female Public Officials: A
Different Voice? Annals ofthe American Academy ofthe Political and Social
Science] 991.

Kingdon, John W. 1989. Congressmen's Voting Decisions. Ann Arbor: University of
Michigan Press.

Kirkpatrick, Jeane J. 1974. Political Women. NY: Basic Books.

Leader. SC. 1977. The Policy Impact of Elected Women Officials. In The Impact of
Electoral Process, edited by L. Maisel and J. Cooper. Beverly Hills
Sage.

Lee. Mei-Hsien. 1989. A Study of Political Roles Played by the R.O.C. Supplementary
Legislators. M.A. Thesis. National Chengchi University, Taipei, Taiwan.

Lij phart, Arend. 1994. Electoral Systems and Party Systems: A Study of Twenty-Seven
Democracies, 1945-1990. New York: Oxford University Press.

Ling. Y.. and A Matsuno. 1992. Women's Struggle for Empowerment in Japan. In

205

I’Vomen Transforming Politics: Worldwide Strategies/or Empowerment. edited by
Bystyzienski. Bloomington: Indiana University Press.

Lu. I-Isiu-Lien Annette. 1994. Women's Liberation: the Taiwanese Experience. In The
Other Taiwan: I 945 to the Present, edited by M. A. Rubinstein. Armonk, N.Y.:
ME. Sharpe.

Matland. Richard E. 1992. District Magnitude's Effect On Female Representation in US.
State Legislatures. Legislative Studies Quarterly 17 (4):469-492.

Mezey. S. 1978. Support for Women's Rights’ Policy. American Political Quarterly
6:485—97.

Muller. C. 1982. Feminism and the New Women in Public Office. Women and Politics 2
(7-21).

Nechemias. Carol. 1994. Democratization and Women's Access to Legislative Seats: The
Soviet Case, 1989-1991. Women and Politics 14 (3):1-18.

Norris. Pippa. 1986. Women in Congress: A Policy Difference. Politics 6:34-40.

Norris. Pippa. 1996. Women PoliticianszTransforming Westminister? Parliamentary
Ajfitir.

Norris. Pippa. and J. Lovenduski. 1995. Political Recruitment: Gender, Race and Class
in the British Parliament. cambridge: Cambridge University Press.

Own-Ying. Tsui-Feng. 1992. The Study of the Legislators' Interpellation on Women's
Issues: a Content Analysis of the Legislative Gazette the 51th to 84th Legislative
Sessions. M.A. thesis. National Chengchi University. Taipei, Taiwan.

Perkins. Jerry, and Diane F owlker. 1980. Opinion Representation Versus Social
Representation; or Why Women Can't Run as Women and Win. American political
Science Review 39:92-103.

Phillips. Anne. 1991. Engendering Democracy. University Park: The Pennsylvania State
University Press.

Phillips. Anne. 1993. Democracy and Difference. Pennsylvania: The Pennsylvania State
University Press.

Phillips. Anne. 1995. The Politics ofPresence. Oxford: Clarendon Press.

Pitkin. Hannah F. 1967. The Concept ofRepresentation. Berkerly: University of
Califonia Press.

Randall. Vicky. 1987. Women and Politics. Chicago: University ofChicago Press.

Reingold. Beth. I992. Concepts of Representation Among Female and Male State
Legislators. Legislative Studies Quarterly. 17:509-537.

Reskin. B. 1988. Bringing the Men Back In: Sex Differentiation and hte Devaluation of
Women's Work. Gender and Society 2:58-81.

Rhode. Deborah L. 1990. Theoretical Perspective on Sexual Difference. In Theoretical
Perspective on Sexual Difference, edited by D. L. Rhode. New Haven: Yale
University Press.

Rubinstein. Murray A. 1994. Introduction: "The Taiwan Miracle". In The Other Taiwan:
1945 to the Present. edited by M. A. Rubinstein. Armonk. N.Y.: ME. Sharpe.

 

Rule. Wilma. 1987. Electoral Systems. Contextual Factors and Women's Opportunity for
Election to Parliament in Twenty-three Democracies. Western Political Quarterly
Septemberz477-98.

Rule. Wilma. and Joseph F. Zimmerman. eds. 1994. Electoral Systems in Cotnparative
Perspective: Their Impact on Women and Minorities: Greenwood Press.

Saint-Germain. Michelle A. 1989. Does Their Difference Make a Difference? The Impact
of Women on Public Policy in the Arizona Legislature. Social Science Quarterly.
70:956-68.

Sapiro, Virginia. 1981. When Are Interests Interesting? The Problem of Political
Representation of Women. American Political Science Review 75 (3):701-716.

Sapiro. Virginia. 1991. Gender Politics. Gendered Politics: the State ofthe Field. In
Political Science .' Looking to the Future. edited by W. Crotty. Illinois:

Northwestern University Press.

Sheng. Shing-Yuan. 1996. Electoral Competition and Legislative Participation. Ph. D.
Dissertation. Political Science. University of Michigan. Ann Arbor.

Skjeie. Hege. 1992. The Rhetoric of Difference: On Women's Inclusion into Political
Elites. Politics and Society 19 (2):233-263.

Soule. J.W., and WE. McGrath. 1977. A Comparative Studies of Male-Female poltiical
Attitudes at Citizen and Elite Levels. In A Portrait of Marginality. edited by N.
Githens and J. Prestage. New York: McKay.

Squires. Judith. 1995. Quotas for Women: Fair Representation? Parliamentary Affairs:

207

Stanwick. K, and K Kleeman. 1983. Women Make a Diflerence. New Brunswick. NJ:
Center for the American Woman and Politics, Rutgers University.

Studlar. Donley. and Susan Welch. 1989. Multimember Districts and the Representation
of Women: Evidence from Britain and the United States. Journal of Politics
(IV'Iay):39I-4l2.

Taagepera. Rein. 1994. Beating the Law of Minority Attrition. In Electoral Systems in
Comparative Perspective: Their Impact on Women and Minorities. edited by W.
Rule and J. F. Zimmerman. Westport: Greenwood Press.

Tamerius. Karin L. 1993. Sex, Gender, and Leadership in the Representation of Women.
In Gender Power, Leadership, and Governance. edited by G. Duerst-Lahti and R.
M. Kelly. Ann Arbor: University of Michigan Press.

Taylor. Charles. 1994. The Politics of Recognition. Edited by A. Gutmann.
Mulliculturalism: Examining the Politics ofRecognition. N. J .: Princeton University

Press.

Thomas. Sue. 1991. The Impact of Women on State Legislative Policies. Journal of
Politics 53:958-76.

Thomas. Sue. 1992. The Effects of Race and Gender on Constituency Service. Western
Political Quarterly 45: 169-80.

Thomas. Sue. 1994. How Women Legislate: Oxford University Press.

T homas. Sue., and Susan. Welch. 1991. The Impact of Gender on Activities and Priorities
of State Legislators. Western Political Science Quarterly 44:445-456.

Tien. Hung-mao. 1996. Election and Taiwan's Democratic Development. In Taiwan '5
Electoral Politics and Democratic Transition: Riding the Third Wave, edited by H.-
m. Tien. Armonk. NY: ME. Sharpe.

Vallance. E. 1979. Women in the House. London: Athlone.

Waylen. Georgina. 1994. Women and Democratization: Conceptualizing Gender
Relations

in Transition Politics. World Politics 46:327-3 54.

Welch. Susan. 1985. Are Women More Liberal than Men in the US Congress. Legislative
Studeis Quarterly 10: 125-34.

Yoder. Janice. 1991. Rethinking Tokenism. Gender and Society 5 (2):178—192.

208

You. Ying-Long. 1989. Gender and Voting Perference. In Women and Political
Participation. edited by S.-L. Liang. Taipei: Awakening Foundation.

Young. Iris Marion. 1990. Justice and the Politics ofDifference. Princeton: Princeton
University Press.