54
Ratchets and See-Saws: Exploring Temporal Patterns in Women’s Political Representation John A. Scherpereel James Madison University [email protected] Suraj Jacob James Madison University [email protected] Melinda Adams James Madison University [email protected] Abstact This paper takes up Darhour and Dahlerup’s (2013) call to determine factors that promote the “sustainable representation of women in politics.” We conceptualize and analyze within-country temporal variation in women’s political representation. We introduce two concepts: the ratchet effect and the see-saw effect. Legislatures are generally characterized by ratchet effects: levels of women’s representation tend to rise over time and experience little backsliding. Cabinets, in contrasts, are often characterized by see-saw effects: representational gains are often followed by declines .After establishing these different patterns, we explore factors that account for variation in the extent of cabinet see-sawing experienced by states. Drawing on case studies of Ireland, Poland, and Japan, we suggest that leadership characteristics, institutional characteristics (particularly the structure of executive-legislative relations), and the structure and focus of women’s movements affect representational trajectories. Paper prepared for presentation at the Joint Sessions of the European Consortium for Political Research, Salamanca, April 2014.

Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

Ratchets and See-Saws: Exploring Temporal Patterns in Women’s Political Representation

John A. Scherpereel James Madison University

[email protected]

Suraj Jacob James Madison University

[email protected]

Melinda Adams James Madison University

[email protected] Abstact This paper takes up Darhour and Dahlerup’s (2013) call to determine factors that promote the “sustainable representation of women in politics.” We conceptualize and analyze within-country temporal variation in women’s political representation. We introduce two concepts: the ratchet effect and the see-saw effect. Legislatures are generally characterized by ratchet effects: levels of women’s representation tend to rise over time and experience little backsliding. Cabinets, in contrasts, are often characterized by see-saw effects: representational gains are often followed by declines .After establishing these different patterns, we explore factors that account for variation in the extent of cabinet see-sawing experienced by states. Drawing on case studies of Ireland, Poland, and Japan, we suggest that leadership characteristics, institutional characteristics (particularly the structure of executive-legislative relations), and the structure and focus of women’s movements affect representational trajectories. Paper prepared for presentation at the Joint Sessions of the European Consortium for Political Research, Salamanca, April 2014.

Page 2: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

1

Levels of women’s political representation in legislature have increased steadily over

time. Figure 1, which plots a long time series of global mean and median values, demonstrates

this trend.1 Even though overall levels of women’s legislative representation still fall far below

parity, the steady forward march in legislative representation is heartening.

To what extent do similar trends characterize cabinets? Because scholars of women and

politics are ultimately interested in women’s relationship to power, and because cabinet positions

often carry more power than legislative positions, this is an important question to answer. Until

recently, however, the lack of global time series data has made it difficult to explore systematic

time trends for cabinets. The time series data collected by Jacob, Scherpereel, and Adams (2014)

and summarized in Figure 2 reveals a trend line that, at first glance, closely resembles the global

trend line for legislatures.

Together, Figures 1 and 2 seem to imply that countries are making slow but steady

progress in both legislatures and cabinets. The paper begins, however, by stressing that

aggregated trend lines can be misleading. We argue, in fact, that legislatures and cabinets are

subject to quite different dynamics. For legislatures, individual countries rarely experience

representational backsliding. Rather, a ratchet effect generally pertains: the gains that women

make in a country’s legislature, however modest, are typically not reversed, and progress is

cumulative.

We argue that there is no equivalent ratcheting in the realm of women’s cabinet

representation. The global trend line for cabinets resembles the global trend line for legislatures,

                                                                                                               1 The data are from Paxton, Green, and Hughes (2003), supplemented by Inter-Parliamentary Union data for later years.

Page 3: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

2

but the cabinet line masks major cross-country variation. The global average is slowly increasing

each year, but individual countries are moving in very different directions. In many countries,

there is a significant increase in women’s cabinet participation one year, followed by a dramatic

decrease in the next year or two. While gains in legislative representation seem to “lock in,”

increases in women’s cabinet participation are often much more precarious. We describe this

phenomenon as a see-saw effect—an asymmetric process in which countries often experience one

or more step back for every two steps forward. Here, we document the see-saw effect and begin

to examine why backsliding (declines in women’s cabinet presence) occurs more often in some

countries than others.

We seek to highlight the difference between the legislative ratchet and the cabinet see-

saw, to explore possible sources of variation in the extent of cabinet see-sawing, and to trace out

methodological and policy implications of the different temporal dynamics that characterize

legislatures and cabinets. Analyzing three cases with different extents of see-sawing—Ireland

(low), Poland (medium), and Japan (high)—we argue that the system of legislative-executive

relations, the preferences of heads of government, and the structures and priorities of the

domestic women's movement help to explain cross-country variation. Leaders who are party

outsiders are more likely to appoint women to cabinet posts. Their ability to appoint women is

affected by their countries’ institutions (e.g., rules regarding who may serve as a minister) and

the strength and focus of women’s movements. Shifts from outsiders to party insiders tend to

promote backsliding, especially in countries where women’s movements are weak and focused

on issues other than women’s political representation.

Page 4: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

3

Global Trends in Women’s Representation

Legislatures: The ratchet effect

For legislatures, trends within individual countries generally mirror global developments;

countries tend to experience slow and steady upward movement. This is somewhat remarkable:

not only do global averages increase year-on-year, but this trend is reflected in the relative

absence of year-on-year reductions in individual countries. To demonstrate this point, we

introduce two simple concepts, backsliding and upward movement. Backsliding for year t is

defined as the change in representation between t-1 and t if the change is negative, and upward

movement is defined as change in representation between t-1 and t if the change is positive.

Figure 3 plots time-series graphs for the percentage of countries with annual backslides and

upward movements in women’s legislative representation. The majority of countries show no

annual variation, which is not surprising given that legislative terms typically last several years.

However, the percentage of countries with annual backslides is particularly low for all years. The

number of countries backsliding in a given year is less than half of the number of countries

experiencing annual upward movements. The figure reveals the presence of a ratchet effect for

most of the world’s legislatures.

A well-established literature directs attention to the factors that affect women’s

legislative representation worldwide (Caul 1999; Matland 1998; Paxton & Hughes 2007, 2010;

Stockemer 2009; Viterna, Fallon, and Beckfield 2008; Wangnerud 2009). Within the legislative

literature, scholars have juxtaposed two pathways—the incremental approach versus the fast-

track approach (Dahlerup and Freidenvall 2005)—through which the proportion of women in

office has increased. In countries characterized by the incremental approach, there is a slow but

Page 5: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

4

steady growth in the proportion of women legislators over time. Nordic countries are the

paradigmatic examples of this model, which is demonstrated in Figure 4. The left graph shows

incremental ratcheting in Nordic legislatures. There are no episodes of dramatic backsliding;

rather, we see periods of stasis interspersed with upward movements. The right graph focuses on

cabinets. Here, by contrast, we see more episodes of backsliding, including a particularly

dramatic backsliding episode in the case of Finland.

The second pathway to increasing women’s legislative representation involves the

adoption of gender quotas, the “fast-track” approach. In fast-track countries (e.g., Senegal), the

adoption of a gender quota leads to a sharp instantaneous increase in the proportion of women

legislators. If implemented properly, legislative gender quotas can also put a floor on women’s

representation, thereby building ratchet effects into formal institutional arrangements (Krook

2009; Franceschet, Krook, and Piscopo 2013).

Interestingly, though, even countries without legislative quotas exhibit year-on-year

increases in women’s legislative representation—and not just the Nordics. This trend is shown in

Figure 5, which focuses on recent years (during the period when quotas have gained momentum

worldwide). The left graph, which focuses on countries that have not adopted quotas, shows that

even those countries have relatively few instances of backsliding. This is true both in absolute

terms and relative to instances of upward movement. In fact, backslides and upward movements

in countries with quotas (right graph) are not dramatically different than in countries without

quotas. This comparison suggests that the factors driving women’s legislative representation—

institutional, social, cultural, and other—are themselves relatively irreversible. It suggests,

furthermore, that the ratchet effect characterizes legislatures in incremental and fast-track

countries alike.

Page 6: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

5

It is neither surprising nor problematic, given this fact, that scholars, activists, and

journalists have tended to focus on upward movement. Activists frequently spotlight countries

when their levels of women’s representation cross a certain threshold (say, 30% or 50%) for the

first time. Within the scholarly literature, there is little discussion of legislative backsliding,

precisely because backsliding is so rare. One exception to the trend comes form post-Soviet

countries and post-communist countries of central and eastern Europe. For the duration of the

communist era, legislatures in these countries were rubber stamps for party policy. As the

process of democratization began and legislatures gained real power, however, women’s

representation dropped off (Paxton and Hughes 2007, 77-80).

Cabinets: the see-saw effect

Backsliding is much more common for women’s cabinet representation than for women’s

legislative representation. To explore backsliding in cabinets, we assembled an original dataset

using information from the CIA’s Chiefs of State and Cabinet Members of Foreign Governments

reports. We have determined the sex of all ministers in the in the CIA reports through Internet

searches, press scrutiny, and correspondence with embassies and country experts. Figure 6 (right

graph) shows, for each year, the percentage of countries with backslides and upward movements.

There are more upward movements than backslides. But what is striking is (a) that the absolute

number of backslides and upward movements is high, (b) that there are many more episodes of

cabinet backslides than legislative backslides, and (c) that there are many more episodes of

cabinet upward movements than legislative upward movements.

To demonstrate the magnitude of upward movements in women’s cabinet representation

and how they are interspersed with episodes of backsliding, we separately aggregate each set of

Page 7: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

6

episodes. Aggregate backsliding is defined as the sum of all year-on-year down backslides.2

Aggregate upward movements are defined analogously. To demonstrate, consider the following

example: In country A, women control the following percentages of cabinet seats for each of the

eleven years between 2000 and 2010: 10, 15, 16, 12, 18, 15, 16, 12, 10, 14, 18. In this case, there

are ten corresponding year-to-year changes: +5, +1, -4, +6, -3, +1, -4, -2, +4, +4. We discard all

cases where year-to-year changes have absolute values of two or lower. This leaves four

instances of upward movement (+5, +6, +4, +4) and three instances of backslide (-4, -3, -4).

Country A’s aggregate upward movement is the sum of its upward movement instance values

(19); its aggregate backslide is the sum of its backslide instance values (11).

Figure 7 presents aggregate backslides and upward movements in 2000-2010 for each

country. The figure brings some interesting points to light. First, the extent of aggregate

backsliding is considerable, averaging as much as 16 percentage points in 2000-2010. Only 16%

of countries have no cabinet backslides in the 2000-2010 period. Second, countries with greater

aggregate backslides tend to have greater aggregate upward movements, as well. The solid line is

the quadratic line of best fit, which shows a strong positive association. These two points are

usefully contrasted with Figure 8, which presents the equivalent figure for legislatures. In the

case of legislatures, the extent of cumulative backsliding is small (averaging less than two

percentage points). Seventy percent of countries have no legislative backslides in this period, and

there appears to be little relationship between aggregate backsliding and upward movements

across countries.

Note that for both cabinets and legislatures (Figures 7 and 8), the extent of aggregate

upward movements is greater than for backslides—for cabinets, averaging 21 percentage points

                                                                                                               2 We use a cutoff of two percentage points to account for the possibility of measurement errors. That is, the aggregates use information only for episodes of backslides and upward movements where annual variation was at least two percentage points.

Page 8: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

7

(compared to 16 for backslides) and for legislatures, averaging 7 percentage points (compared to

less than two for backslides). This is why both institutions exhibit a similar global upward trend

line for the 2000-2010 period.

Figures 9 and 10 provide two additional pieces of information that serve as robustness

checks to the emerging argument that backslides and upward movements are much larger in

cabinets than legislatures. Figure 9 presents aggregate cabinet backslides and upward movements

in 2000-2010 for countries whose average cabinet representation was close to the global mean.

The fact that it resembles Figure 7 in terms of magnitudes suggests that large global backslides

and upward movements are not being driven by outliers. Figure 10 presents the percentage of

countries with backslides in each year, where representation is defined in terms of absolute

numbers—number of women in the cabinet—rather than percentage. On average, 18% of

countries experienced a net decrease of one woman from the cabinet, 8% experienced a net

decrease of two, and 3% experienced a net decrease of three.

This discussion establishes that women’s cabinet representation is fundamentally

different than women’s legislative representation. Cabinet representation is characterized by see-

saws—interspersed upward movements and backslides. Legislative representation is

characterized by ratchets. See-saws are particularly important, we posit, because backslides

erode progress towards gender equality. In this sense, we pick up on Darhour and Dahlerup’s

(2013, 133) concept of “sustainable representation of women in politics,” which they define as

“durable, substantial numerical political representation of women, freed of the risk of immediate

major backlash.” We suggest that the absence—or at least the reduction—of backslides can serve

as a measure of the sustainability of women’s representation.

Page 9: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

8

The fact that individual country developments look very different in the executive branch

than in the legislative branch calls for closer analysis of the factors that lead to backsliding.

Which Factors Might Affect Cabinet See-Sawing?

Worldwide, cabinets are much less likely to experience ratcheting than legislatures. But, as

the analysis above has shown, there is significant cross-country variation in the extent of cabinet

see-sawing. Some countries experience dramatic shifts between upward movement and

backsliding, while others experience steadier temporal paths. While there are few studies of

women’s cabinet representation over time, scholars have begun to identify factors that affect the

likelihood that a cabinet at any particular moment will contain more or less women. These

studies offer a useful theoretical point of departure and help to generate intuitions about why

some more upward movement and backsliding than others. Three sets of possible explanations

emerge from the literature—one on political institutions, the second on leader characteristics, and

the third on characteristics of women’s movements.

Institutional characteristics

Leaders are often constrained by rules (constitutional or otherwise) governing who can

serve as a minister. Most parliamentary systems permit cabinet ministers hold legislative

mandates; others require them to be legislators. In contrast, many presidential systems expressly

prohibit ministers from serving simultaneously in the legislature. Where ministers must be drawn

from the legislature (in part or in full), the pool of potential women ministers is likely to be more

limited, thereby limiting both upward movements and backslides. This distinction is similar to

Page 10: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

9

Davis’s (1997) intuition that leaders in states with specialist systems that prohibit ministers from

being MPs (e.g., the United States) are able to recruit ministerial appointees from a variety of

professional arenas where qualified women may be more plentiful, while leaders in states with

generalist systems (e.g., the United Kingdom) are limited to recruiting ministers from the

legislature.

It is worth noting, however, that Davis concentrates on the West European context, where

economies are more differentiated and pools of professional women in the economy at large are

generally deeper than in the rest of the world. Nonetheless, evidence from developing countries

also suggests that generalist cabinets may impede women’s cabinet presence. In Ghana, for

example, John Evans Atta Mills announced a commitment, during the 2008 presidential

campaign, to achieve at least 40% women in government. After winning the election, however,

Mills reneged on this promise, arguing that since only five women from his party won legislative

seats and the majority of cabinet appointees must be MPs according to Ghanaian law, there was

not a sufficient pool of eligible women to keep his commitment. In short, legislative-executive

relations shape the pool of potential women ministers, increasing or limiting the opportunities for

upward movement or backsliding. Specialized systems may decrease the likelihood of large

upward movements and backsliding.

Leader characteristics

The process of government formation depends critically upon the values and dispositions

of the head of government. Leaders who identify as feminists, for example, should be more

likely to appoint women to cabinet posts than non-feminist leaders. After appointing a parity

cabinet in 2004, for example, Spanish Prime Minister José Luis Rodríguez Zapatero commented:

Page 11: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

10

“I am not just anti-machismo, I’m a feminist” (Williams 2011). Zapatero maintained a near-

parity government throughout his time in office. After the 2008 elections, women held eight out

of 17 cabinet posts. During subsequent cabinet reshuffles, the proportion of women dropped

slightly below 50%. When Mariano Rajoy succeeded Zapatero, the proportion of women in the

cabinet declined, suggesting that a change from a feminist to a non-feminist head of government

can lead to backsliding.

Leaders’ ties with their parties may also influence their likelihood of appointing women.

Siavelis and Morgenstern (2008) have created a typology of executives that includes party

insiders, party adherents, and free-wheeling independents. Party insiders are predicted to appoint

ministers with “strong party credentials” (32). Given the underrepresentation of women among

party leaders, this fact may decrease the likelihood of women gaining cabinet posts. The cabinets

of free-wheeling independents, on the other hand, are “likely to be ad hoc, made up of outsiders,

and based primarily on the personal networks of support” (34). These characteristics may create

openings for more women to gain access to the cabinet. In an analysis of Latin American

presidents’ appointments, Escobar-Lemmon and Taylor-Robinson (2008) find that party insiders

appoint fewer women to their cabinets and that free-wheeling independents appoint more women

to ministerial posts. They speculate that the “promotion of women might be yet another way for

[a free-wheeling independent] to exhibit his or her independent style and to show that his or her

administration is a break from the past” (349).

The literature also suggests that leaders’ political ideology will affect the likelihood that

they will appoint women to the cabinet. Specifically, leaders that identify with the left will be

more likely to include women in their cabinets, due to left parties’ commitment to egalitarianism

(Reynolds 1999; Escobar-Lemmon and Taylor-Robinson 2008). Reynolds (1999) finds that the

Page 12: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

11

composition of the coalition (e.g., left leadership or inclusion of a left party) affects women’s

inclusion more than left parties’ share of the vote. Escobar-Lemmon and Taylor-Robinson

(2008) find that party loyalists and party adherents from left parties are more likely to appoint

women to their cabinets than party loyalists and party adherents from right parties. This

relationship, however, does not hold for free-wheeling independents.

One might expect, given studies of leadership characteristics, that shifts from feminists to

non-feminist, from left-dominated governance to right-dominate governance, and from free-

wheeling independents to party insiders would coincide with episodes of backsliding.

Characteristics of women’s movements

In their global study of women’s cabinet participation, Bauer and Tremblay (2011, 184)

argue that the gains in women’s cabinet presence would not be possible without women’s

organizations that promote women’s political representation. Countries with strong women’s

movements mobilized around women’s political representation should have less backsliding than

those without active women’s movements. Women’s movements, on this view, may provide a

floor for women’s inclusion that prevents large declines in women’s representation. By

organizing protests, publicly criticizing parties and leaders, and imposing other costs on leaders

and parties that fail to include a sufficient number of women in government, women’s

movements can decrease the likelihood of backslides.

Page 13: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

12

Institutional Characteristics: A Preliminary Probe

While the possible effects of leader characteristics and women’s movements are best

observed through fine-grained case studies (see below), we can conduct a preliminary test to

determine whether certain institutional characteristics might be associated with cabinet see-saws.

To carry out one such test, we code countries according to whether (a) legislators cannot be

ministers, (b) legislators can be ministers, and (c) ministers must be legislators. Our coding,

which is based on Fish and Kroenig (2009), allowed us to code 78% of the countries in our

dataset of 200. Of these countries, slightly less than half (44%) fit into category (a), 21% in

category (b), and 13% in category (c). Figure 11 shows aggregate upward movements and

backslides averaged for each category. Backsliding and upward movement are both greatest

where heads of government must look outside of the legislature (the first pair of bars in Figure

10). Both types of movement are less, on average, in countries where a part of the cabinet is

appointed from amongst the set of legislators (the second pair of bars). Movements are still less

in countries that limit cabinet seats exclusively to legislators. Interestingly, countries that we

could not code have the least see-sawing – these are typically small countries that do not make it

to cross-national databases like Polity.3

A second institutional factor that may drive upward movements and backslides is cabinet

size. Larger cabinets may allow chief executives to distribute patronage and minimize the risk of

extra-constitutional threats to their rule (Arriola 2009), but, we reason, each increase in the size

of the cabinet marginally decreases the prestige of the cabinet as an institution. The cost of

                                                                                                               3  These are Albania, Algeria, Andorra, Antigua & Barbuda, Aruba, Bahamas, Barbados, Belize, Bermuda, Brunei, Burma, Cape Verde, Djibouti, Dominica, Equatorial Guinea, Ethiopia, Grenada, Iceland, Kiribati, Kosovo, Liechtenstein, Luxembourg, Maldives, Malta, Marshall Islands, Micronesia, Monaco, Montenegro, Nauru, Palau, Saint Kitts & Nevis, Saint Lucia, Saint Vincent & Grenadines, Samoa, San Marino, Sao Tome & Principe, Serbia & Montenegro, Seychelles, Solomon Islands, Somalia, Suriname, Tonga, Tuvalu, Vanuatu, Yugoslavia.  

Page 14: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

13

replacing a dismissed woman minister with another woman may therefore be higher in a state

with a small cabinet than it is in a state with a large cabinet. That is, smaller cabinets may make

it more difficult for chief executives to replace departing women with new women.

To gauge whether larger cabinet size might reduce backslides, we plot the two variables in

Figure 11 (we use our original database to calculate cabinet size). The first part of Figure 12 does

show a negative relationship – that is, on average, larger sized cabinets have fewer aggregate

backslides in 2000-2010. Similarly, on average, larger-sized cabinets have less aggregate upward

movements, too. The difference, therefore, which gives the net change in women’s cabinet

representation, does not have much relationship with cabinet size. However, there is considerable

noise in the scatter plot, suggesting that much of the variation in both aggregate backslide and

aggregate upward movement is left unexplained by accounting for cabinet size alone. The size of

the legislature, similarly, does not appear to be much related to aggregate upward movements

and backslides, as shown in Figure 13.

Case studies

To generate deeper insights into the possible effects of institutional, leadership, and

women’s movement characteristics, we consider three relatively similar cases with different

dependent variable outcomes. As Figure 14 demonstrates, Ireland has experienced few upward

movements or backslides between 2000 and 2014. In the early 2000s, Poland experienced

significant see-sawing but, since 2007, has held relatively steady. Japan has experienced

dramatic see-sawing over the period observed.

Page 15: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

14

Ireland

Viewed from a certain (acute) angle, Ireland is a global trendsetter in women’s executive

leadership; for 21 of the last 25 years, the country’s president (the head of state) has been a

woman, and Ireland is the only country in the world to have had one woman president (Mary

McAleese, 1997-2011) replace another (Mary Robinson, 1990-1997). Robinson and McAleese

increased the visibility and political relevance of the Irish presidency, and both women, despite

different ideological backgrounds, have advocated throughout their careers on behalf of women’s

rights.

As Figure 14 makes clear, however, Ireland is not a leader in the executive offices that

matter most. In the cabinet, the country’s recent temporal trajectory involves stagnation.

Although two women—Mary Harney (1997-2006) and Mary Coughlan (2008-2011)—have

recently served as deputy prime minister (Tánaiste), overall levels of women’s presence have

remained flat. In addition, the few women who have achieved cabinet office in recent years have

failed to ascend to high-power posts. Women ministers in the last two cabinets have been kept

out of high-power, high-prestige posts, reverting to gendered positions such as education, health

and children, and social protection. Cultural and institutional characteristics, in addition to

factors related to the structure of the Irish women’s movement, stand out as potentially important

sources of these trends. Traditional Catholicism is an always-present background variable in

modern Ireland, and members of the country’s church hierarchy have generally emphasized

women’s roles as mothers and caregivers. Within this context, women have faced significant

hurdles with respect to labor force incorporation, equal pay, reproductive rights, and gender-

based violence.

Page 16: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

15

The historical and institutional features of Irish democracy have made it difficult for Irish

women to address these issues from within state structures. Unlike most West European states,

Ireland’s party system has not been characterized by a strong and balanced left-right split. The

division between the republic’s two main parties—Fianna Fáil and Fine Gael—can be traced

back to disputes about the Anglo-Irish relationship (instead of disputes related to capitalist

industrialization). As Mair and Weeks (2005, 136) noted, the two main parties “tend to converge

around the centre of the ideological spectrum, often crossing sides between centre-left and

centre-right, or occupying both simultaneously.” Their respective European party affiliations—

Fianna Fáil and Fine Gael belong, respectively, to the Alliance of Liberals and Democrats for

Europe (ALDE) and the European People’s Party (EPP)—suggest that the parties’ self-

conceptions remain right of center.

Throughout the period of independence, other parties have gained seats in parliament

and, since 1989, Ireland has been ruled exclusively by coalition governments. According to

Gallagher et al.’s calculations (2006, 232-241) Ireland’s left (including the long-standing Labour

Party and other, smaller left parties) was weaker than the left in any other West European state

over the course of the post-war period. In the most recent (2011) general elections, which took

place in the wake of the financial crisis, Labour took advantage of the governing parties’ (Fianna

Fáil and the Greens) deep unpopularity and, for the first time in its history, finished second in an

election. Labour eventually formed a two-party coalition with Enda Kenny’s Fine Gael. Time

will tell whether this development results in a deeper restructuring of the party system or a

fundamental fortification of the Irish left. Historically, nonetheless, the record is clear: Ireland’s

left has been quite weak.

Page 17: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

16

The Irish constitution establishes a particularly strict fusion of executive and legislative

powers. It requires that the prime minister (Taoiseach), deputy prime minister (Tánaiste), and

minister of finance be selected from, and serve simultaneously as members of, the country’s

lower legislative chamber (the Dáil Éireann). It places the same stipulation on all other ministers,

with the slight exception that a maximum of two of those ministers may be members of the upper

chamber (the Seanad Éireann). A prime minister wishing to increase women’s representation in

the cabinet is, therefore, constitutionally restricted in a way that most presidents (and, indeed,

many prime ministers in parliamentary systems) are not – effectively, he or she may only choose

women from a pool of TDs (Teachtaí Dála, or Dáil members). For this reason, cabinet dynamics

depend critically upon electoral system design and parties’ systems of candidate selection.

Ireland uses a PR-single-transferrable vote system to select TDs. The country is divided

into multiple multi-member districts, which return between three and five TDs apiece. Voters

may use ballots to indicate ordinal preferences, marking a “1” adjacent to their favored

candidate, a “2” next to their second favorite candidate, etc. One notable feature of Irish STV is

that a party may select more than one candidate to place on a ballot. Depending on district

characteristics, such a decision represents a certain risk—it pays off if voters select multiple

candidates from the same party and backfires of voters split votes between a single party’s

candidates.

Candidate selection processes vary by party, but, traditionally, selection processes have

been decentralized and animated by a localist ethos. The established parties (Fianna Fáil, Fine

Gael, and Labour) have placed a premium on candidate visibility within electoral districts; this

norm has often militated against the selection of women candidates. From the late 1970s

onwards, newer/smaller parties and Fine Gael spearheaded efforts to promote women as

Page 18: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

17

parliamentary candidates. The defunct Progressive Democrats and currently reeling Greens, for

example, adopted 40% candidate quotas in the early 1990s, and Galligan (1993, 157) showcases

the “discovery of gender by the traditional parties from the 1977 general elections onward.”

Developments at the candidate selection phase, however, have not had a strong effect on

levels of women’s legislative representation. One might have expected Labour—as the country’s

most stable force on the left and that party that chose Mary Robinson as its presidential

candidate—to have prioritized the selection of women. In the legislative arena, however,

Labour’s strategy has often been driven by a minimal-loss strategy (Galligan 1993); motivated

by fears of splitting Labour votes, the party has often placed only one name on each

constituency’s ballot. The selection, in this case, mirrors the gendered strategic dynamics (e.g.,

“if we can only choose one candidate, we will choose a man, since some voters are sexist”) that

have often been observed in single-member district electoral systems (Matland 2002). Thus, the

configuration of factors discussed so far—a traditional culture, a weak left, an electoral system

that personalizes PR, and a system of legislative-executive relations that restrict cabinet

eligibility to members of a single legislative chamber, help to stack the deck against aspiring

women ministers.

These structural barriers aside, Irish prime ministers still have significant discretion to

increase and decrease women’s cabinet shares. Why have they not taken more frequent

advantage of this discretion? As noted above, coalition government has been the norm in Ireland

since 1989. Perhaps, as Krook and O’Brien (2012) have suggested, coalition governance has

worked against women in the cabinet. The intuition here is that there is a relatively finite number

of cabinet positions; the greater the number of parties in a cabinet, the fewer party members any

one party can appoint to the cabinet. Because party leaders tend to be men, it is unlikely for

Page 19: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

18

women to take on cabinet seats. In Ireland, however, this intuition is not supported. The 2000s,

for example, were dominated by a Fianna Fáil-Progressive Democrat coalition. The Progressive

Democrats were a small liberal party that collapsed at the end of the decade. For the better part

of the 2000s, their one and only cabinet post was held by the party leader—Mary Harney.

Despite having many more seats to distribute, senior coalition partner Fianna Fáil never

distributed more than two ministerial dossiers to women. The Fine Gael-Labour coalition that

has been in power since 2011 has been characterized by rather similar dynamics. The junior

partner (Labour) has consistently allocated one of its relatively few (five) cabinet seats to a

woman, while the senior partner (Fine Gael), despite its surfeit of seats, has only assigned one

seat to a woman.

It is also worth noting that Ireland has lacked an avowedly feminist Taoiseach and that all

Irish prime ministers over the last three decades have been “creatures of the system.” The

premiership has bounced back and forth between Fianna Fáil and Fine Gael, but the political

biographies of the men who have served as Taoiseach are remarkably similar. Before taking on

the premiership, the six men who have served as prime minister since 1979 have possessed, on

average, over 23 years of service as a TD (Albert Reynolds, with fifteen years as a TD, had the

least experience; Enda Kenny, with 36 years, had the most). All six prime ministers had served

in multiple ministerial roles before becoming prime minister, and all had survived years of intra-

party politicking within their respective parties (Fine Gael for John Bruton and Enda Kenny;

Fianna Fáil for Charles Haughey, Albert Reynolds, Bertie Ahern, and Brian Cowen).

The clubbiness of the Irish political system might seem ripe for challenges from below,

e.g., from the Irish women’s movement. The scholarship on the movement is large and diverse,

covering multiple historical periods. Coulter (1993, 59) proposes that, over the grand sweep of

Page 20: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

19

the nineteenth and twentieth centuries, Irish women’s groups and movements have positioned

themselves as alternatives to the state, rather than as means toward achieving state power:

The great movements which have involved women in this country over the past two hundred years have all developed outside the institutions of the existing state, and have usually been opposed to them. . . . women have been accustomed to finding alternative ways of organizing and networking. They have fixed their eyes, not on capturing positions within the existing institutions, but on creating different institutions, which would be more responsive to the needs of the majority of people, particularly women and children. If a strong strain of anti-political and locally based organizing has characterized Irish

women’s mobilization, scholars (Connolly 2002; Coulter 1993; Galligan 1998; Reilly 2007;

Smyth 1988) generally agree that the early 1970s witnessed more movement crystallization,

followed by a return to more decentralized battles. From the mid-1970s onward, fragmentation,

issue-oriented mobilization, and intra-organizational campaigns have again become predominant.

This fragmentation has not necessarily led to disappointment. Indeed, the awakening of parties to

issues of women’s representation, discussed above, owed much to the strategic activities of

women’s campaigns within the various parties in the late 1970s and early 1980s. Diverse

organizations picked up on the global “women’s rights as human rights” framework, interpreting

and deploying it in diverse contexts, and occasionally achieving goals (Reilly 2007). The

traditional decentralization and extra-institutional nature of Irish women’s mobilizing, however,

arguably worked against a concerted and highly organized local push, in the wake of Beijing and

other international developments in recent decades, for increased women’s representation in

parliament and the cabinet.

Ireland has a strong history of women’s mobilizing, and women have ascended to the

presidency and the deputy premiership. Party leaders, over the past decades, have clearly felt

some pressure to appoint women to cabinet posts and to avoid dramatic drops in women’s

Page 21: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

20

cabinet presence. At the same time, institutional factors and the decentralized and policy-

centered (as opposed to representation-centered) nature of the Irish movement continue to work

against upward movement in women’s cabinet representation.

Poland

Women’s cabinet representation in Poland rose significantly between 2000 and 2014. In

January 2000, Poland had no women ministers. By January 2011, this number had risen to 29.4%

(five out of seventeen) and has remained at a similar level in recent years. Currently, four of the

country’s seventeen ministers (23.5%) are women. The Polish line in Figure 14 see-saws in the

early 2000s before climbing and plateauing from 2008 onward.

Which developments might explain this trajectory? In terms of context, Poland resembles

Ireland in numerous ways. Poles and Irish citizens are sometimes seen as “the strongest

Catholics left in Europe.” Catholicism has played a central role in both countries’ nationalist

movements, and national Catholic hierarchies have been active (and, on occasion, actively

hostile to gender equality initiatives) in both states. In partisan terms, both states are competitive

democracies that have been dominated, at least over the last decade, by the political right.

Poland’s post-communist left enjoyed significant popularity in the 1990s and early 2000s (see

below) but has struggled to reestablish itself following devastating corruption scandals. Since

2005, Poland has witnessed a pattern of vigorous intra-right conflict that resembles, without

directly mimicking, the Fianna Fáil-Fine Gael dynamic in Ireland. In Poland, the two major

protagonists of this millennium’s politics have been Civic Platform (a liberal-conservative party

that has ruled in coalition with the Polish People’s Party since 2007) and Law and Justice (a

Page 22: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

21

national-conservative party that led an ill-fated coalition between 2005 and 2007 and has been

the major opposition party ever since).

In terms of executive-legislative relations, however, Polish prime ministers are enabled

and constrained in different ways than their Irish counterparts. On one hand, the Polish head of

state (the president) enjoys more political power than the Irish head of state. While Poland is

often classified as a “semi-presidential regime,” it is more accurately characterized as a premier-

presidential regime (Shugart and Carey 1992)—a regime in which the president takes legislative

partisan balances and negotiations into account, nominates a prime minister for parliament’s

confirmation, and lacks the power to dismiss the PM or individual ministers once they have

achieved parliament’s vote of confidence. Discord between presidents and prime ministers often

occurs in such systems, particularly (but not exclusively) in moments of cohabitation (Protsyk

2006). Regardless of the party identities of the office holders, however, Polish presidents have

tended to leave the details of cabinet formation—including personnel selection and portfolio

allocation—to prime ministers. Other constitutional provisions reinforce the prime minister’s

relative freedom, particularly Article 103, which allows prime ministers to legislators and “non-

legislators” to cabinet posts. Unlike Irish prime ministers, whose pool of potential women

ministers is limited to the universe of women parliamentarians, Polish prime ministers can (and

often do) choose women (and men) who lack legislative mandates. Of the nine women who have

served as ministers between 2010 and 2014, for example, four have been sitting MPs; the other

five have come from outside of parliament.

While the Polish left has languished for the last decade, an understanding of gendered

cabinet dynamics since 2000 must begin with the observations that the early 2000s belonged to

the Polish center-left. The Democratic Left Alliance (SLD)/Labor Union coalition, dominated by

Page 23: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

22

the SLD and headed by Leszek Miller, easily won the September 2001 elections, garnering 41%

of the popular vote and nearly half of the seats in Poland’s lower chamber (the Sejm). Maciej

Płażyński’s Civic Platform, which was contesting parliamentary elections for the first time, came

in second place with (only) 12.7% of the vote, and the only other parties to receive seats were

four conservative and nationalist parties (including Law and Justice, which was also contesting

its first national election). The fragmented right-wing parties were divided by personality and

ideology and had a hard time finding common ground. As a result, Miller initially led a minority

government that was very secure and faced few viable threats. For this reason, he may have felt

comfortable appointing a cabinet (in 2001) with only two female ministers.

The SLD government’s margin of comfort quickly narrowed, though. Over the course

2002 and 2003, it pushed through a series of unpopular austerity measures. It supported the

unpopular US-led invasion of Iraq. It found itself embroiled in a deeply unpopular corruption

scandal that revolved around film producer Lew Rywin. Its growing distance from voters (and

the increasing popularity of various right-wing alternatives) may have encouraged Miller to

increase the number of female ministers in hopes of attracting female voters. Right-wing parties,

internally divided and confident that their electoral fates would improve as time passed, elected

not to push for snap elections. As a result, upon Miller’s early resignation in May 2004, a kind of

caretaker center-left government led by the SLD’s Marek Belka took office. Interestingly, this

government had zero female ministers. The SLD seems to have become so unpopular at this time

that Belka went into deep preservation mode, trying to minimize losses rather than expanding the

party’s broader electoral appeal. In other words, Belka may have appointed more established

(male) ministers who would placate the SLD’s hard core of strong supporters, despite the SLD’s

shrinking overall attractiveness.

Page 24: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

23

In the years that followed, the SLD sustained massive losses, and robust competition

between the liberal right (Civic Platform) and the nationalist right (Law and Justice) intensified.

The September 2005 elections represented a major shift in Polish politics and were completely

dominated by different strands of broadly right-wing groupings. Law and Justice, headed by

Jaroslaw Kaczyński, won 27% of the votes, followed by Civic Platform (24.1%), the nationalist

Self-Defense Party (11.4%), the SLD (11.3%), the conservative/nationalist League of Polish

Families (8%), and the agro-conservative Polish People’s Party (7%). Law and Justice had run

on a socially conservative platform, campaigning against abortion rights and defending

traditional family values. Kaczyński assembled a coalition with Self-Defense and the League of

Polish Families. One of the resulting government’s first legislative achievements was a pro-

natalist bill that established payments to women for each new child they bore. The government

clearly clung to a traditional vision of gender relations but seemed eager to use female

appointments to parry the opposition’s accusations of sexism and backwardness. The Law and

Justice-led government appointed two female ministers in 2005 and two more in 2006. What is

more, of the four ministries that women controlled in January 2007, two (finance and foreign

affairs) were high-prestige postings.

As the leaders of the coalition’s two junior partners committed various gaffes and the

liberal opposition (Civic Platform) continued to gain steam, the Sejm voted to dissolve itself in

2007. In the resulting early elections, Civic Platform (with 41.5% of the votes) emerged

victorious and formed a minimum-winning center-right coalition with the Polish People’s Party

(which had received 8.9% of the vote). Since 2007, this coalition has lasted. Opposition from the

left continues to be weak. But opposition from the conservative/nationalist right has been

significant. While Civic Platform won the 2010 presidential elections and became the first senior

Page 25: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

24

governing party in post-communist Poland to be reelected in 2013, Poland’s party system has

become quite competitive, despite the lack of a strong center-left alternative.

While party competition seems to be an important factor driving Poland’s trend line, the

Polish women’s movement—which has emerged as a strong, unified, and centralized political

force over the course of the last five years—has also encouraged prime ministers not to

backslide. Since the 1990s, a diverse and dedicated set of Polish feminists has agitated for

women’s concerns. These have not been fringe movements; rather, various Polish newspapers

and magazines with wide distributions (e.g., Gazeta Wyborcza) have featured regular

contributions from leading academics and activists. Also, in 2007, best-selling novelist Manuela

Gretkowska founded the Women’s Party, which, despite its eventually dire showing at the polls

(0.28%), was one of the most commented-upon phenomena of the campaign (due largely, but not

exclusively, to a campaign poster featuring nude party leaders). For each of the past five years, a

highly visible organization, the Congress of Polish Women, has met to discuss the status of

Polish women in general and, more specifically, the incorporation of Polish women into political

decision-making positions. The Congress’s annual plenary session is large (estimated annual

attendance for the last five years has been 4000, 4000, 7000, 10000, 8500), and is generally

preceded by a series of regional congresses. The Congress is a diverse affair, bringing together

women leaders from the worlds of business, politics, NGOs, the academy, labor unions, and

elsewhere, and it is notable for being a wide partisan tent. The Congress’s plenary meetings have

become major media events, and the group’s leaders have been granted numerous high-level

meetings with Poland’s presidents and prime ministers.

One of the Congress’s first demands was a parity law, which would have require 50% of

the names of all party lists for parliamentary, European Parliament, and local elections to be

Page 26: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

25

women. The Congress did not meet this goal, but politicians from all parties were forced to take

it seriously. A parity bill was eventually submitted to parliament, and, after a long debate that

also involved consideration of electoral systems in both national legislative chambers and a

range of sub-national legislatures, Poland passed a law on legislative quotas for elections to the

Sejm. As of 2011, at least 35% of the names on all electoral lists must be men, and at least 35%

of the names on all electoral lists must be women. While many women’s advocates had hoped

for a zipper system that would prohibit parties from relegating women to the bottom of lists, the

quota seems to have mattered in the 2011 election. The percentage of women candidates doubled

in 2011 vs. the previous election (2007), and women’s share of parliamentary seats increased

moderately from 20% to 23.7%.

How do these developments relate to women’s representation in the cabinet? There has

been no law on cabinet quotas, but the prominence of the legislative quota campaign and the

broader visibility of women’s issues has clearly encouraged all political parties, regardless of

their ideological persuasion, to consider gender balance in public positions. Druciarek et al.

(2012, 10) considered the extent to which the quota debate affected parties’ approaches to Polish

local government elections (which were not covered by the quota law) and found that while

quotas “were not applied to the local government elections, [they] undoubtedly influenced the

tone of the public debate that preceded the 2010 voting. The issue of women’s participation in

politics became one of the key issues discussed during the electoral campaign.” Gender issues

also featured prominently in the 2011 electoral campaign and its aftermath. While the lead story

of that election was the success of the incumbent (Civic Platform-Polish People’s Party)

government, an important secondary story was the third-place finish by Palikot’s Movement, an

ideologically ambiguous secular and youth-oriented party led by former Civic Movement

Page 27: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

26

member Janusz Palikot. Palikot’s Movement placed more women on its lists (43.9%) than any

party besides the SLD, and two of its most visible successful candidates were Robert Biedroń

(Poland’s first openly gay MP) and Anna Grodzka (Europe’s first openly transsexual MP).

Despite the weakness of the Polish left and the traditionalism of Polish culture, the

discretion left to prime ministers, the competitiveness of the party system, and, especially, the

vigor and common purpose of civil societal mobilization have helped to keep gender issues on

the agenda and to sustain women’s progress in cabinet representation.

Japan

As Figure 14 demonstrates, Japan experienced dramatic upward movement in women’s

cabinet representation in the early 2000s. Within three years, however, Japan had experience

significant backsliding, and, over the last few years, very few Japanese women have ascended to

high ministerial positions. The structure of party politics and weaknesses of the women’s

movement help to explain Japan’s see-sawing pattern.

The Japanese constitution ensures a moderately tight fusion between executive and

legislative branches but grants the prime minister some discretion to make appointments from

beyond the Diet (Japan’s bicameral legislature). Specifically, Article 68 of the constitution

requires at least half of all cabinet ministers, including the prime minister, to be a member of

either the House of Representative or the House of Councilors. Theoretically, then, a prime

minister may appoint a significant number of individuals without legislative mandates to his or

her cabinet.4 An amendment passed in 2001, however, capped the total number of cabinet

                                                                                                               4 Constitutionally, the Japanese parliament selects the prime minister via votes in both chambers of the National Diet. If the two houses select different individuals and the dispute is not resolved within ten days, the House of Representatives’ choice is sent to the Emperor (Secretariat 2014). The Emperor appoints the individual selected by

Page 28: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

27

members (not including the prime minister) at fourteen, which could be expanded, under

ambiguous conditions of “special necessity,” to seventeen. Thus, the constitutional framework

affords Japanese prime ministers a degree of appointment freedom somewhere between the

highly restrictive Irish case and the more permissive Polish case.

The Liberal Democratic Party (LDP), which has dominated Japanese politics for the bulk

of the post-war period, has exercised its power within this basic constitutional context. Because a

majority of cabinet members must come from the Diet, it is necessary to mention the significant

barriers faced by women seeking legislative seats. Anyone seeking to represent the LDP in the

Diet, for example, must build a party constituency (jiban), gain publicity and endorsements

(kanban), and raise money (kaban). Women (especially those without connections to husbands,

fathers, or other male relatives who are or have been politicians) face disproportionately high

hurdles in gaining access to these three resources (Ogai 2001; Reed, McElwain, and Shimizu

2009; Stockwin 2008).

While these hurdles have often discouraged women from seeking legislative seats, they

have been further handicapped by the structure of the LDP and links between the LDP and the

state. For most of the post-war period, the LDP has dominated the Diet, and the LDP party

president has served simultaneously as prime minister (Kabashima and Steel 2007). The LDP

constitution limits party presidents to two consecutive three-year terms, so LDP prime ministers

have traditionally faced de facto six-year limits. Before 2001, the party constitution incentivized

networking with LDP Diet members.5 As a result of the party-state relationship and the internal

                                                                                                                                                                                                                                                                                                                                                                     the Diet (or, in the rare case of Diet division, the individual selected by the lower chamber) as prime minister. Once appointed, the prime minister selects line ministers. 5 To be nominated as a candidate for the party presidency, a prospective party leader had to receive endorsements from thirty Diet members (not including himself). All candidates who received at least thirty endorsements then went before the party convention. The convention brought together (a) all current LDP members from the House of Representatives, (b) all current LDP members from the House of Councilors, and (c) party delegates from LDP chapters in each of the country’s 47 prefectures. Before 2001, each prefecture had only one vote in the party

Page 29: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

28

LDP rules, party presidential (and, thus, prime ministerial) candidates were especially responsive

to LDP Diet members’ concerns and tended to discount provincial concerns (Reed, McElwain,

and Shimizu 2009).

LDP Diet members, aware of their king-making powers, organized themselves into strict,

hierarchical factions and voted in blocs at the party convention. Faction leaders provided their

juniors with leadership positions of increasing prestige, based on seniority, in exchange for

support in intra-party leadership elections. Not surprisingly, the structure of Japan’s legislative

electoral system affected the structure of LDP factions. Specifically, the country’s multi-member

district/single, non-transferrable vote system (MMD-SNTV) limited faction size and required

factions to form coalitions with each other in order to elect party presidents. In general, to win

the LDP presidency (and, thus, the premiership), a candidate required the support of three or

more factions. Altogether, the system rewarded long-serving party loyalists and worked against

party mavericks or upstarts.

In the early 1990s, however, the LDP’s grip on power weakened. Numerous high-level

party members resigned from the party and established new parties. The LDP sustained heavy

losses in the 1993 general election, and, for the first time since 1955, a coalition that did not

include the LDP (an eight-party alliance led by Morihiro Hosokawa) was formed. Hosokawa’s

government fell apart within a year. Its only notable accomplishment was an overhaul of the

election system that replaced MMD-SNTV with a joint PR/SMD-P system modeled loosely on

Germany. Hosokawa’s government was replaced by an LDP-Socialist Party coalition in 1994,

but the LDP remained shaken by the 1993 defeat. In an attempt to recover, the party—facing

significant pressure from prefectural branches and the media—reformed its presidential selection

                                                                                                                                                                                                                                                                                                                                                                     presidential selection process. This placed prefectural delegates at a distinct disadvantage vis-à-vis the 300-400 Diet standing members, since each Diet member possessed one vote apiece.

Page 30: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

29

process in a way that gave more strength to prefectural branches (Reed, McElwain, and Shimizu

2009).6 This intra-LDP shake-up provided the backdrop of the dramatic upward movement in

Japanese women’s cabinet representation in the early 2000s.

The new system provided an opportunity for Junichiro Koizumi to ascend to the positions

of LDP boss and prime minister. While Koizumi is often remembered as a maverick, he was not

an LDP outsider. His grandfather and his father had served in the Diet. Koizumi himself had

been a member of the House of Representatives since 1972 and had held three separate cabinet

posts between 1988 and 1998. After the shocks of the early 1990s, however, Koizumi staked out

ground as a leader of the party’s reformers, and he took advantage of the new system of party

elections to emerge as party leader in 2001 (Reed, McElwain, and Shimizu 2009; Stockwin

2008).

Koizumi’s championing of women’s representation was an integral component of his

initial appeal. By supporting women—a group that had always been dramatically

underrepresented in the party and the state—Koizumi could attract new supporters to the party

and burnish his credentials as a maverick (Gelb 2003; Kobayashi 2004). When Koizumi took on

the premiership in 2001, he selected women to lead five of eleven ministerial dossiers—foreign

affairs; justice; environment; education, culture, sport, science, and technology; and land,

infrastructure, and transport. Two of the five women (Atsuke Toyama at Education and Yoriko

Kawaguchi at Environment) did not hold Diet seats at the time of their selection, furthering the

                                                                                                               6 Three particular party reforms paved the way for the spike. First, the number of Diet endorsements needed to become a presidential candidate decreased from thirty to twenty, effectively lowering the bar for candidacy. Second, the prefectures’ weight in the selection process was tripled; instead of one vote, each prefectural chapter would now have three votes. Third, prefectural branches would be permitted (a) to decide if they wanted to administer an open primary at the prefectural level and (b) to determine how to allocate their three votes. In the event, most prefectural branches elected to use a winner-take-all system. Diet members would be restricted to endorsing candidates that had been selected by the prefectures. Diet members were also allowed to see all prefectural results prior to voting, thereby reducing the influence of faction bosses on individual Diet members’ behavior (Reed, McElwain, and Shimizu 2009; Rosenbluth and Thies 2010; Stockwin 2008).

Page 31: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

30

impression of “freshness.” Makiko Tanaka, the Minister of Foreign Affairs, had an insider’s

pedigree—she was a serving MP, and her father, Kakuei Tanaka, was a two-time prime minister

and LDP scion—but was still seen as a firebrand within the party, not unlike Koizumi himself.

Tensions between Koizumi and Tanaka mounted however, and Koizumi replaced her in 2002.

Despite Tanaka’s departure, Koizumi seems to have remained at least partially wedded to

the strategy of promoting women; he promoted environment minister Kawaguchi to the foreign

affairs portfolio but filled the vacant environmental post with a man, Hiroshi Oki. At numerous

points throughout his premiership (2001-2006), Koizumi championed women leaders. In the run

up to the 2005 elections in which LDP would triumph, for example, he endorsed women

opponents, stylized as “female assassins,” over the LDP incumbents who had opposed his postal

reform package (Reed, McElwain, and Shimizu 2009).

In many ways, however, the Koizumi case highlights the distinction between party

outsiders and feminist heads of government. Koizumi was certainly the former; but it is not clear

that he was also the latter. Unlike José Luis Zapatero in Spain, for example, Koizumi never

identified himself as a feminist. One searches in vain for evidence of a strong personal

commitment to gender equality in Koizumi’s pre-premiership biography, and Koizumi’s

commitment to women’s empowerment within in the cabinet clearly waned as his term as prime

minister developed. After the LDP’s electoral success in the 2003 elections, for example,

Koizumi appointed only two women to the cabinet (Yuriko Koike at Environment and, again,

Yoriko Kawaguchi at Foreign Affairs), and, by time he stepped down in 2006 (due to the LDP

party term limit), only Koike was the only woman minister remaining.

While Koizumi helped to restabilize the LDP and led the party in three successful

election campaigns (two for the House of Representatives, one for the House of Councilors), his

Page 32: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

31

three LDP successors as prime minister—Shinzo Abe (2006-2007), Yasuo Fukuda (2007-2008),

and Taro Aso (August 2008-September 2008)—decreased the number of women in the cabinet.

Abe, Fukuda, and Aso all enjoyed the traditional support of multiple LDP factions (Krauss and

Pekkanen 2010) and relied for their respect ascents on Diet members rather than prefectural

delegates. For these reasons, and in the absence of sustained pressure from outside the party (see

below), they appointed cabinets with no women leading portfolios.

The Democratic Party of Japan—the LDP’s major oppositional rival—politicized the

LDP’s gender backsliding in the 2009 House of Representatives elections. The DPJ has

traditionally been more attuned to women’s and gender issues than the LDP (Reed, McElwain,

and Shimizu 2009), and Ichiro Ozawa, the DPJ’s head campaign strategist in 2009, intentionally

ran 34 female candidates against older, male LDP candidates. The effort proved successful, with

26 of the 34 being elected into office. While “Ozawa’s Girls” only constitute one-fifth of the

DPJ’s gains in the election, they still made headlines. From its early days in power, however, the

DPJ struggled to maintain party loyalty. None of the three DPJ prime ministers—Yukio

Hatoyama (September 2009-June 2010), Naoto Kan (June 2010-September 2011), and

Yoshihiko Noda (September 2011-December 2012)—evidenced of deep commitment to

women’s representation, and “Ozawa’s Girls” did not make strong inroads into the cabinet. All

three DPJ prime ministers “awarded” women with two junior minister positions (these positions

are not counted in the data displayed in Table *, which are restricted to full ministers); all of the

DPJ’s full ministerial positions were distributed to men.

To this point, the discussion has stressed that at particular moments—e.g., the moment of

LDP crisis in the 1990s and the 2009 electoral campaign—actors within political parties

(Koizumi, Ozawa) have sought political advantage through the strategic promotion of women.

Page 33: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

32

To what extent have they been assisted or constrained by independent women’s groups? While

patterns of Japanese women’s mobilization have changed over time (Gelb 2003), most observers

agree that the women’s organizing has not been strongly feminist in content and that the

movement has tended to orient itself toward broader social or public issues rather than issues

related to gendered power dynamics or unequal representation (Fujimura-Fanselow 2011; Liddle

and Nakajima 2000; Matsui 1990; Murase 2006). In the nineteenth century, women played

significant roles in Japan’s Popular Rights Movement, which ultimately helped to achieve the

ratification of the Meiji Constitution (1889). While that document created a national assembly, it

severely restricted the franchise, excluding women entirely and restricting the suffrage to one

percent of the male population. Meiji-era rules went further in restricting women’s activity and

political mobilization with the creation, via the 1898 civil code, of the ie system. This system

defined the role of women as wives and mothers, transferred women’s property to their

husbands, and gave the head of household (usually the eldest male) unquestioned authority over

all household members. In addition, Article 5 of 1900’s Police and Security Regulations banned

women from joining political organizations, attending meetings with political speeches and

lectures, and organizing any such meetings (Fujimura-Fanselow 2011; Liddle and Nakajima

2000; Patessio 2011; Stockwin 2008).

These legal foundations, which would prove temporally sticky, did not succeed in

eliminating women’s mobilizing, but they steered such mobilizing into decentralized, apolitical

fujinkai (women’s groups). During the Taisho period (1912-1926), a germinal first-wave national

women’s movement came to life, which re-emerged after WWII. In 1945, Japanese women

finally gained the suffrage, and in 1947 the ie system was eliminated.

Page 34: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

33

As the postwar period developed, though, Japanese women’s organizing came to bear a

closer resemblance to the localist and depoliticized fujinkai than to the nationally visible,

politically motivated suffrage and anti-ie movements. While a nascent Second Wave movement

took root in the early 1970s, it never achieved the concentration or success that some Western

movements reached (Liddle and Nakajima 2000; Matsui 1990; Rosenbluth and Gies 2010). In

her contemporary study of 889 women’s organizations at the local, prefectural, and national

levels, Murase (2006) finds that just 31% of such organizations deal with women’s issues, such

as domestic violence and women retaining their surnames after marriage. Only a third of that

31% “explicitly embrace gender equality as their goal.” Overall, Murase concludes that there is

no central issue or policy focus that unifies Japanese women’s groups. Women’s political

representation, for example, is not the star cause around which women’s efforts rotate, in the

manner that suffrage or the elimination of ie had been for earlier generations. The diversity and

depoliticization of the Japanese women’s movement make it difficult for the movement to

provide a “floor” below which levels of women’s cabinet presence may not sink.

It is in this environment—a democracy where women are enfranchised and frequently

turn out to vote, where party environments (as in the LDP after 1993) provide institutional

opportunities for mavericks, and where the women’s movement is decentralized, depoliticized,

and diverse—that Japanese see-sawing has developed.

Conclusions

Worldwide, levels of women’s representation are rising in both legislatures and cabinets.

We have emphasized, though, that the similar global trend lines mask important differences in

Page 35: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

34

intra-country dynamics. In legislatures, most countries experience ratchet effects—once levels of

women’s representation reach a certain level, they rarely drop much below that level. Cabinets,

on the other hand, often experience see-sawing; backsliding frequently follows upward

movement. Because of this fact, women’s relationship to the heights of public power is often

tenuous. The notion that levels of women’s representation are growing may be encouraging, but

scholars and activists must bear in mind that cabinet gains are often (but not always) transitory

and that gains in year t may well be wiped out in year t+1 or t+2.

Our basic statistical probes suggest that there may be some relationship between certain

institutional features—most notably, the ability of a prime minister to appoint ministers from

outside of the legislature—and temporal trajectories in women’s cabinet representation. The

three cases, in addition, suggest that certain combinations of leaders, institutions, and women’s

movement characteristics are more likely to limit backsliding than others. Ireland’s institutional

set-up does not lend itself to upward movements in women’s cabinet presence. The constitution

prohibits prime ministers from appointing non-MPs to cabinet posts, and the country’s electoral

system (STV) poses significant obstacles to women seeking legislative office. In addition, while

the Irish women’s movement has managed to unite disparate streams at particular historical

moments, its dominant mode of organizing has involved anti-political, issue-based, and often

localist concerns. This decentralized nature has paid off in certain ways – advocates used to

mobilizing in local venues have carried out significant women’s promotion campaigns, for

example, within particular Irish political parties. These dynamics, in turn, have allowed Ireland

to hold steady at approximately 20% over the course of the current millennium. But they have

not promoted significant upward movement.

Page 36: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

35

The Irish case contrasts with that of Poland, which, after years of see-sawing, has more

recently witnessed upward movement and “lock-in.” Poland’s institutional environment is more

permissive than Ireland’s. The constitution neither restricts cabinet posts to the universe of sitting

MPs nor caps total cabinet size, and the legislative electoral system (party-list PR, recently

supplemented by a 35% quota) opens pathways to women’s candidacy. Poland’s women’s

movement, furthermore, has been more cohesive and broader than Ireland’s or Japan’s, and the

movement, since the late 2000s, has consistently emphasized women’s political representation.

These conditions, in addition to related developments in the party system (e.g., the rise of

Palikot’s Movement), have kept issues of women’s representation visible to leaders from all

political parties and seem to have placed a limit on backsliding.

In Japan, the institutional barriers that worked against women’s cabinet representation

have decreased somewhat in recent years. The party system has become more competitive. The

rules within the LDP have been rewritten in a way that generates more opportunities for party

outsiders. While the majority of the seats in the House of Representatives are now determined on

the basis of single-member district/plurality elections, a significant number of seats (180 out of

480) are allocated via party lists, and parties are allowed to list particular individuals

simultaneously on both SMD-P ballots and party lists. These institutional changes open spaces

for “mavericks” and generate opportunities for aspiring women politicians, but other institutional

features (e.g., the upper limit on the number of cabinet members and the limitations on

appointments from outside the Diet) continue to constrain Japanese prime ministers. Beyond the

institutional setting, Japanese see-sawing has been enabled by the absence of a cohesive

movement that prioritizes representational issues. Some male politicians have taken strategic

advantage of women’s status as “untainted outsiders.” But the country has had no devoted

Page 37: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

36

feminist leader, and the likelihood of such a leader emerging is low, given the fragmentation of

the movement and the lack of an explicitly representational movement agenda.

While the paper’s empirical sections have not directly tested the potential relevance of

partisan (left-right) leadership for gender see-sawing and ratcheting, it is worth noting, in

conclusion, that Ireland, Poland, and Japan are all “weak left” cases. Many of the headline-

grabbing examples of women’s cabinet empowerment—including Zapatero’s Spain and Renzi’s

Italy—have involved parties of the left with (in Zapatero’s case) avowedly feminist leaders.

Together, Ireland, Poland, and Japan suggest that left leaders are not necessary for (a) significant

upward movement in women’s cabinet representation (Japan) or (b) maintenance of women’s

representational gains (Ireland and Poland). Future research might subject the notions that left-

wing parties promote upward movement and/or prevent backsliding to deeper scrutiny.

In the meantime, scholars and activists should continue to pay close attention to issues of

women’s cabinet representation, appreciating that while cabinet dynamics may be related to

developments in the legislature, the cabinet is generally (a) more powerful and (b) more volatile

than the legislature. Regardless of institutional features, a party leadership committed to

women’s representation and a cohesive women’s movement that prioritizes representation can

help to limit backsliding and promote more equal representational outcomes.

Page 38: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

37

Figure 1: Women’s Representation in the World’s Legislatures

05

1015

20

perc

enta

ge p

oint

s

1940 1960 1980 2000year

mean medianFitted values

Page 39: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

38

Figure 2: Women’s Representation in the World’s Cabinets

05

1015

perc

enta

ge p

oint

s

1979 1989 1999 2009year

mean medianFitted values

Page 40: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

39

Figure 3: Backslides and Upward Movements in Women’s Legislative Representation

010

2030

40

perc

enta

ge o

f cou

ntrie

s

1940 1960 1980 2000year

increase decreasedecrease (more than 2 percentage points)

Page 41: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

40

Figure 4: Nordic Countries

010

2030

4050

60

1979 1989 1999 2009year

Denmark FinlandNorway Sweden

legislature

010

2030

4050

60

1979 1989 1999 2009year

Denmark FinlandNorway Sweden

cabinet

Page 42: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

41

Figure 5: Backslides and Upward Movements in Women’s Legislative Representation, for Countries with and without Quotas

010

2030

40

1995 2000 2005 2010year

increase decreasedecrease (more than 2 percentage points)

countries without quotas

010

2030

40

1995 2000 2005 2010year

increase decreasedecrease (more than 2 percentage points)

countries with quotas

Page 43: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

42

Figure 6: Legislatures vs. Cabinets: Backslides and Upward Movements

010

2030

40

2000 2005 2010year

increase decreasedecrease (more than 2 percentage points)

legislature

010

2030

40

2000 2005 2010year

increase decreasedecrease (more than 2 percentage points)

cabinet

Page 44: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

43

Figure 7: Aggregate Backslides and Upward Movements in Women’s Cabinet Representation

AFG

ALB

DZA

ADO

AGO

ATG

ARG

ARMABW

AUS

AUT

AZE

BHS

BHR

BGD

BRB

BLR

BEL

BLZ BEN

BMU

BTN

BOL

BIH

BWABRA

BRN

BGR

BFA

MMR

BDI

KHM

CMR

CANCPV

CAF

TCD

CHL

CHN

COLCOM

ZAR

COG

CRI

CIV

HRV

CUB

CYP

CZE

DNK

DJI

DMA

DOM

ECU

EGY

SLV

GNQ ERI

EST

ETH

FJI

FIN

FRA

GAB

GMB

GEO

DEU

GHA

GRC

GRD

GTM

GIN

GNB

GUY

HTI

HND

HUN

ISL

IND IDN

IRN

IRQ

IRL

ISR

ITA

JAM

JPN

JOR

KAZ

KEN

KIRPRK

KOR

KSV KWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBY

LIE

LTU

LUX

MKD

MDG

MWI

MYS MDV

MLI

MLT

MHL

MRT

MUS

MEX

FSM

MDA

MCO

MNG

MNE

MAR

MOZNAM

NPL

NLD

NZL

NIC

NER

NGA

NOR

OMN

PAK PLW

PAN

PNG

PRY

PER

PHL

POL

PRT

QAT

ROM

RUS

RWA

KNA

LCA

VCT

WSM

SMR

STP

SAU

SEN

SRB

SYC

SLE

SGP

SVK

SVN

SLB SOM

ZAF

ESP

LKA

SDNSUR

SWZ

SWE

CHE

SYR

TWN

TJK

TZA

THA

TMP

TGO

TON

TTO

TUN TUR

TKM

UGA

UKR

ARE

GBR

USA

URY

UZB

VUT

VEN

VNM

YEM

SRB

ZMB

ZWE

020

4060

aggr

egat

e ba

cksl

idin

g in

200

0-20

10 (%

)

0 20 40 60 80aggregate upward movements in 2000-2010 (%)

Page 45: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

44

Figure 8: Aggregate Backslides and Upward Movements in Women’s Legislative Representation

AFG ALBDZA

ADO

AGOATG

ARG

ARMABW

AUS

AUTAZE

BHS

BHR

BGD

BRB

BLRBEL

BLZ

BENBMU

BTN

BOL BIH

BWA

BRABRN

BGR

BFAMMR BDIKHMCMRCAN CPVCAF TCDCHLCHNCOLCOM

ZARCOG

CRICIV

HRV

CUBCYPCZEDNK DJI

DMA

DOM ECUEGY

SLV

GNQ

ERI ESTETH FJIFIN FRA GAB

GMBGEO

DEU

GHA

GRC

GRD

GTM GIN

GNB

GUYHTI

HND

HUN

ISL

IND IDNIRN

IRQ

IRL ISR ITAJAM JPNJORKAZKEN

KIR

PRK

KORKSV KWT KGZLAO

LVA

LBN LSOLBRLBY LIE

LTU

LUX MKD

MDG

MWIMYS

MDV

MLI

MLTMHL MRT MUS

MEX

FSM MDAMCO

MNG

MNE

MARMOZ

NAM

NPL

NLD

NZL

NIC

NERNGA NOR

OMN

PAKPLW

PAN

PNG PRY PER

PHL

POL PRTQAT ROMRUS RWA

KNA

LCA

VCT

WSM

SMR

STPSAU SENSRB

SYC

SLE SGP

SVK

SVN

SLBSOM

ZAFESPLKA SDN

SUR

SWZ

SWE

CHESYRTWN TJK TZATHATMP TGOTON TTOTUNTUR

TKM

UGA

UKR

AREGBRUSA URY UZBVUT

VEN

VNMYEM SRBZMB

ZWE

05

1015

aggr

egat

e ba

cksl

idin

g in

200

0-20

10 (%

)

0 10 20 30 40aggregate upward movements in 2000-2010 (%)

Page 46: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

45

Figure 9: Aggregate Backslides and Upward Movements in Women’s Cabinet Representation for Countries Near the Global Mean

AGO

BGD

BOL

BFA

CAF

COG

CIV

DOM

EST

GABGRD

GNB

GUY

HTI

HND

IRL

ITA

JAM

JPN

KAZ

KSV

MWI

MLT

MEX

NGA

PRT

ROM

VCT

SRB

SUR

TZATGO

ZMB

020

4060

aggr

egat

e ba

cksl

idin

g in

200

0-20

10 (%

)

0 20 40 60 80aggregate upward movements in 2000-2010 (%)

Page 47: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

46

Figure 10: Aggregate Backslides in Women’s Cabinet Representation, Absolute Values

05

1015

2025

perc

enta

ge o

f cou

ntrie

s

2000 2005 2010year

decrease decrease (> 1)decrease (> 2)

Page 48: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

47

Figure 11: Relationships between Legislatures and Cabinets

 

   

05

1015

2025

leg. cannot be min. leg. may be min. min. has to be leg. pathway unclear

upward movements backsliding

Page 49: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

48

AFG

ALB

DZA

ADO

AGOATG

ARG

ARMABW

AUS

AUT

AZE

BHS

BHR

BGD

BRB

BLR

BEL

BLZBEN

BMU

BTN

BOL

BIH

BWABRA

BRN

BGR

BFA

MMR

BDI

KHM

CMR

CANCPV

CAF

TCD

CHL

CHN

COLCOM

ZAR

COG

CRI

CIV

HRV

CUB

CYP

CZE

DNK

DJI

DMA

DOM

ECU

EGY

SLV

GNQERI

EST

ETH

FJI

FIN

FRA

GAB

GMBGEO

DEU

GHA

GRC

GRD

GTM

GIN

GNB

GUY

HTI

HND

HUN

ISL

INDIDN

IRN

IRQ

IRL

ISRITA

JAM

JPN

JORKAZ

KEN

KIR PRK

KOR

KSVKWT

KGZ

LAOLVA

LBN

LSO

LBR

LBY

LIE

LTU

LUX

MKDMDG

MWIMYSMDV

MLI

MLT

MHL

MRT

MUS

MEX

FSM

MDA

MCO

MNG

MNEMAR

MOZNAM

NPL

NLD

NZL

NICNER

NGA

NOR

OMN

PAKPLW

PAN

PNG

PRY

PER

PHL

POL

PRT

QAT

ROM

RUS

RWA

KNA

LCA

VCT

WSM

SMR

STP

SAU

SEN

SRB

SYC

SLE

SGP

SVK

SVN

SLB SOMZAF

ESP

LKA

SDNSUR

SWZ

SWECHE

SYR

TWN

TJK

TZA

THA

TMP

TGOTON

TTO

TUNTUR

TKM

UGA

UKR

ARE

GBR

USA

URY

UZB

VUTVEN

VNM

YEM

SRB

ZMB

ZWE

020

4060

agg.

bac

kslid

ing

(%)

0 20 40 60

backsliding

AFG

ALBDZA

ADO AGOATG

ARG

ARM

ABW

AUS

AUT

AZE

BHSBHR

BGD

BRB BLR

BEL

BLZBEN

BMUBTN

BOL

BIH

BWA

BRA

BRN

BGR

BFA

MMR

BDI

KHM

CMRCAN

CPV

CAF

TCD

CHL

CHN

COL

COM ZAR

COG

CRI

CIV

HRV

CUB

CYP

CZE DNK

DJI

DMADOM

ECU

EGY

SLV

GNQ

ERI

EST

ETHFJI

FIN

FRA

GAB

GMB

GEO

DEU

GHA

GRC

GRD

GTM

GIN

GNB

GUY

HTI

HND

HUN

ISL

IND

IDN

IRN

IRQ

IRL

ISRITA

JAM

JPN

JOR

KAZ KEN

KIR

PRK

KOR

KSV

KWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBY

LIE

LTU

LUX

MKD

MDG

MWI

MYSMDV MLI

MLT

MHLMRT

MUS

MEX

FSM

MDA

MCO MNGMNE

MARMOZ

NAM

NPL

NLD

NZL

NIC

NER

NGA

NOR

OMN

PAK

PLW

PAN

PNG

PRY

PER

PHL

POL

PRT

QAT

ROM

RUS

RWA

KNA

LCA

VCTWSMSMR

STP

SAU

SEN

SRB

SYC SLESGP

SVK

SVN

SLB

SOM

ZAF

ESP

LKA

SDNSUR

SWZ

SWE

CHE

SYR

TWN

TJKTZA

THA

TMP

TGO

TON

TTO

TUN

TUR TKMUGAUKR

ARE

GBR

USA

URY

UZB

VUT

VEN

VNM

YEM

SRB

ZMB

ZWE

020

4060

80

agg.

upw

ard

mov

emen

ts(%

)0 20 40 60

upward movements

AFG

ALB

DZAADO

AGOATGARG ARM

ABW AUS

AUT

AZE

BHS

BHR

BGDBRB

BLRBEL

BLZBEN

BMUBTN

BOL

BIHBWA

BRA

BRN

BGRBFA

MMR

BDI

KHMCMR

CAN

CPV

CAF TCD

CHL

CHN

COL

COMZAR

COG

CRI

CIV

HRV

CUB

CYP

CZEDNK

DJI

DMADOM

ECU

EGY

SLV

GNQERI

ESTETH

FJI

FINFRA GABGMB

GEODEU

GHAGRC

GRD

GTM

GIN

GNB

GUY

HTI

HNDHUN

ISL IND

IDNIRN

IRQIRL ISRITA

JAMJPN

JORKAZ

KEN

KIR

PRKKORKSV

KWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBY

LIE

LTU

LUXMKD

MDG

MWI

MYSMDV

MLI

MLT

MHL

MRT MUSMEX

FSM

MDA

MCO

MNG

MNE

MAR

MOZ

NAMNPL

NLDNZL

NIC

NERNGA

NOR OMN

PAK

PLW

PAN PNGPRY PER PHL

POLPRT

QAT

ROM

RUSRWAKNA

LCA

VCTWSM

SMR

STP

SAU

SEN

SRB

SYC

SLE

SGP

SVK

SVN

SLB

SOM

ZAF

ESP

LKA

SDNSUR

SWZSWE

CHE

SYRTWN

TJK

TZA

THA

TMP

TGO

TONTTO

TUN

TUR

TKM

UGA

UKR

ARE

GBRUSA

URY

UZB

VUT

VEN

VNM YEMSRBZMB

ZWE

-40

-20

020

4060

net c

hang

e (%

)

0 20 40 60average cabinet size (%)

net change

AFGALB

DZA

ADO

AGOATG

ARG

ARM

ABW

AUS

AUT

AZE

BHS

BHR

BGD

BRBBLR

BEL

BLZBEN

BMU

BTN

BOL

BIH

BWA

BRA

BRN

BGR

BFA

MMR

BDI

KHM

CMRCAN

CPV

CAF

TCD

CHL

CHN

COL

COMZAR

COG

CRI

CIV

HRV

CUB

CYP

CZEDNK

DJI

DMA

DOM

ECU

EGY

SLV

GNQERI

EST

ETH

FJI

FIN

FRA

GAB

GMB GEO

DEU

GHA

GRC

GRD

GTM

GIN

GNB

GUY

HTI

HND

HUN

ISL

INDIDN

IRN

IRQ

IRL

ISRITA

JAM

JPN

JORKAZ

KEN

KIRPRK

KOR

KSVKWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBYLIE

LTU

LUX

MKDMDGMWI

MYSMDV

MLIMLT

MHL

MRT

MUS

MEX

FSM

MDA

MCO

MNG

MNE

MARMOZ

NAM

NPL

NLD

NZL

NIC

NER

NGA

NOR

OMN

PAK

PLW

PAN

PNG

PRY

PER

PHL

POL

PRT

QAT

ROM

RUS

RWA

KNALCAVCT

WSM

SMR

STP

SAU

SENSRB

SYC

SLE

SGP

SVKSVN

SLBSOM

ZAF

ESP

LKA

SDNSUR

SWZ

SWE

CHE

SYR

TWN

TJK TZATHA

TMPTGO

TON

TTO

TUNTUR

TKMUGA

UKRARE

GBR

USA

URY

UZB

VUT

VEN

VNM

YEM

SRB

ZMB

ZWE

050

100

150

back

slid

ing

+ up

war

d m

ovem

ents

(%)

0 20 40 60average cabinet size (%)

tot. seesaw

Figure 12: See-saws and Size of the Cabinet

 

Page 50: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

49

AFG ALB DZAAGOATG

ARG

ARM

AUS

AUTAZEBHR

BGD

BLR BELBEN

BTN

BOLBIH

BWA

BRA

BGR

BFAMMR BDIKHM CMR CANCAF TCDCHL COLCOM

ZARCOG

CRI CIV

HRV

CUBCZEDNKDJI DOMECU EGY

SLV

GNQ

ERIESTFJI FIN FRAGAB

GMBGEO

DEU

GHA

GRCGTMGIN

GNB

GUYHTI

HND

HUN

INDIDNIRN

IRQ

IRLISR ITAJAM JPNJORKAZ KEN

PRK

KORKWT KGZ LAO

LVA

LBNLSOLBRLBY

LTU

MKDMDG

MWIMYS

MLI

MHL MRTMUS

MEX

MDAMCO

MNG

MARMOZ

NAM

NPL

NLD

NZL

NIC

NER NGANOR

OMN

PAK

PAN

PNGPRY PER

PHL

POLPRTQAT ROM RUSRWA

LCASMR

SAU SENSRBSRBSLESGP

SVK

SVN

SLBSOM

ZAFESPLKASDNSWZ

SWE

CHE SYRTWNTJK TZA THATGOTTO TUN TUR

TKM

UGA

UKR

ARE GBRUSAURY UZB

VEN

VNMYEMZMB

ZWE

05

1015

agg.

bac

kslid

ing

(%)

0 200 400 600 800

backsliding

AFG

ALB

DZA

AGO

ATGARGARMAUS

AUT

AZE

BHR

BGD

BLR

BEL

BEN

BTN

BOLBIH

BWA

BRA

BGR

BFA

MMR

BDI

KHM

CMR

CANCAF

TCDCHL

COLCOM

ZAR

COG

CRI

CIV

HRV

CUB

CZE

DNK

DJI

DOM

ECU

EGY

SLV

GNQ

ERI

EST

FJIFIN

FRAGAB

GMB

GEO

DEUGHA

GRC

GTM

GIN

GNB

GUY

HTI HND HUN IND

IDN

IRN IRQIRL

ISR

ITA

JAM

JPNJORKAZKEN

PRK

KORKWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBY

LTU

MKD

MDG

MWI

MYSMLIMHL

MRT

MUS MEX

MDA

MCOMNG

MARMOZ

NAM

NPL

NLD

NZL

NICNER

NGANOR

OMN PAK

PAN

PNG

PRYPER

PHL

POLPRT

QAT

ROMRUS

RWA

LCASMR

SAU

SENSRB

SRB

SLE

SGP

SVK

SVN

SLBSOM

ZAFESP

LKA

SDN

SWZ

SWE

CHE

SYRTWN

TJK

TZA

THATGO

TTO TUN

TUR

TKM

UGA

UKR

ARE

GBR

USA

URY

UZB

VEN

VNMYEMZMB

ZWE

010

2030

40

agg.

upw

ard

mov

emen

ts(%

)

0 200 400 600 800

upward movements

AFG

ALB

DZA

AGO

ATG

ARGARM

AUS

AUT

AZEBHR

BGD

BLR

BEL

BEN

BTN

BOLBIH

BWA

BRA

BGR

BFA

MMR

BDI

KHMCMR

CANCAFTCDCHLCOLCOM ZAR

COG

CRI

CIV

HRV CUB

CZE

DNK

DJI

DOM

ECU

EGYSLV

GNQ

ERIEST

FJI FINFRAGABGMB

GEO DEUGHA

GRC

GTM

GIN

GNB

GUY

HTI

HND

HUNIND

IDN

IRN

IRQ

IRL

ISR

ITA

JAM

JPNJORKAZKEN

PRK

KORKWT

KGZ

LAOLVALBN

LSO

LBR

LBY LTU

MKD

MDG

MWI

MYSMLI

MHL

MRT

MUSMEX

MDA

MCO

MNG

MARMOZ

NAM

NPL

NLDNZL

NICNER

NGANOR

OMNPAKPANPNG

PRYPER

PHLPOL

PRT

QAT

ROM RUS

RWA

LCASMR

SAU

SENSRB

SRB

SLE

SGP

SVKSVN

SLBSOM

ZAFESP

LKA

SDN

SWZ

SWE

CHE

SYRTWN

TJK

TZA

THATGO

TTO TUN

TUR

TKM

UGA

UKR

ARE

GBRUSA

URY

UZB

VEN

VNMYEMZMBZWE

-20

020

40

net c

hang

e (%

)

0 200 400 600 800average legislature size (%)

net change

AFG

ALB

DZA

AGO

ATGARG

ARMAUSAUT

AZE

BHR

BGD

BLR

BEL

BEN

BTNBOL

BIH

BWA

BRA

BGR

BFA

MMR

BDI

KHM

CMR

CANCAF

TCDCHL

COLCOM

ZAR

COG

CRI

CIV

HRV

CUB

CZE

DNK

DJI

DOM

ECU

EGY

SLV

GNQ

ERI

ESTFJI

FINFRA

GAB

GMB

GEO

DEU

GHA

GRC

GTM

GIN

GNB

GUY

HTI

HND

HUNIND

IDN

IRN

IRQ

IRL

ISR

ITA

JAM

JPNJORKAZKEN PRK

KORKWT

KGZ

LAO

LVA

LBN

LSO

LBR

LBY

LTUMKD

MDG

MWI

MYSMLI

MHL

MRT

MUSMEX

MDA

MCO

MNG MARMOZ

NAM

NPL

NLD

NZL

NIC

NER

NGANOR

OMNPAK

PAN

PNG

PRYPER

PHL

POLPRT

QAT

ROMRUS

RWA

LCASMR

SAU

SENSRB

SRB

SLE

SGP

SVK

SVNSLBSOM

ZAFESP

LKA

SDN

SWZSWE

CHE

SYRTWN

TJK

TZA

THATGO

TTO TUN

TUR

TKMUGA

UKR

ARE

GBR

USA

URY

UZB

VEN

VNMYEMZMB

ZWE

010

2030

40

back

slid

ing

+ up

war

d m

ovem

ents

(%)

0 200 400 600 800average legislature size (%)

tot. seesaw

Figure 13: See-saws and Size of Legislature

   

Page 51: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

50

Figure 14: Women’s Political Representation in Ireland, Poland, and Japan

010

2030

4050

2000 2005 2010year

Ireland

010

2030

4050

2000 2005 2010year

Poland010

203040

50

2000 2005 2010year

cabinet legislature

Japan

Page 52: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

51

Works Cited

Aleman, Jose and David Yang. 2011. “A Duration Analysis of Democratic Transitions and Authoritarian Backslides.” Comparative Political Studies 44(1): 1-29.

Arriola, Leo. 2009. “Patronage and Political Stability in Africa.” Comparative Political Studies 42(1): 1339-1362.

Bauer, Gretchen and Manon Tremblay, eds. 2011. Women in Executive Power: A Global Overview. New York: Routledge.

Caul, Miki. 1999. “Women’s Representation in Parliament: The Role of Political Parties.” Party Politics 5(1): 79-98.

Connolly, Linda. 2002. The Irish Women’s Movement: From Revolution to Devolution. London: Palgrave.

Coulter, Carol. 1993. The Hidden Tradition: Feminism, Women and Nationalism in Ireland, Cork: Cork University Press.

Dahlerup, Drude and Lenita Freidenvall. 2005. “Quotas as a “Fast Track” to Equal Representation for Women: Why Scandinavia is No Longer the Model.” International Feminist Journal of Politics 7(1): 26-48.

Darhour, Hanane and Drude Dahlerup. 2013. “Sustainable Representation of Women through Gender Quotas: A Decade’s Experience in Morocco.” Women’s Studies International Forum 41(2): 132-142.

Davis, Rebecca. 1997. Women and Power In Parliamentary Democracies: Cabinet Appointments in Western Europe, 1968-1992. Lincoln: University of Nebraska Press.

Druciarek, Malgorzata, Malgorzata Fuszara, Aleksandra Nizynska, and Jaroslaw Zbieranek. 2012. Women on the Polish Political Scene. Warsaw: Institute of Public Affairs.

Epstein, David, Robert Bates, Jack Goldstone, Ida Kristensen and Shannon O’Halloran. 2006. “Democratic Transitions.” American Journal of Political Science 50(3): 551-69.

Escobar-Lemmon, Maria and Michelle M. Taylor-Robinson. 2008. “How Do Candidate Recruitment and Selection Processes Affect the Representation of Women?” in Peter M. Siavelis and Scott Morgenstern, eds., Pathways to Power: Political Recruitment and Candidate Selection in Latin America. State College, PA: Penn State University Press: 345-368.

Fish, M. Steven and Matthew Kroenig. 2009. The Handbook of National Legislatures: A Global Survey New York: Cambridge University Press.

Franceschet, Susan, Mona Lena Krook, and Jennifer M. Piscopo, eds. 2012. The Impact of Gender Quotas. New York: Oxford University Press.

Fujimura-Fanselow, Kumiko, ed. 2011. Transforming Japan: How Feminism and Diversity Are Making a Difference. New York: The Feminist Press at the City University of New York.

Gallagher, Micahel, Michael Laver, and Peter Mair. 2006. Representative Government in Modern Europe: Institutions, Parties, and Governments, 4 ed. New York: McGraw Hill.

Galligan, Yvonne. 1998. Women and Politics in Contemporary Ireland: From the Margins to the Mainstream. London: Pinter.

Galligan, Yvonne. 1993. “Party Politics and Gender in the Republic of Ireland.” In Joni Lovenduski and Pippa Norris, eds.,Gender and Party Politics. London: Sage: 147-168.

Gelb, Joyce. 2003. Gender Policies in Japan and the United States: Comparing Women's Movements, Rights, and Politics. Basingstoke: Palgrave Macmillan. Jacob, Suraj, John Scherpereel, and Melinda Adams 2014. “Gender Norms and Women’s

Page 53: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

52

Political Representation: A Global Analysis of Cabinets, 1979-2009.” Governance: An International Journal of Policy, Administration, and Institutions 27(2): 321-345.

Kabashima, Ikuo, and Gill Steel. 2007. “How Junichiro Koizumi seized the leadership of Japan's Liberal Democratic Party.” Japanese Journal of Political Science 8(1): 95-114.

Kaufman, Joyce P., and Kristen P. Williams. 2010. Women and War: Gender Identity and Activism in Times of Conflict, Sterling, VA: Kumarian Press, 2010.

Kobayashi, Yoshie. 2004. “Has the Closed Door Opened for Women? The Appointment of Women Ministers in Japan.” PS: Political Science and Politics 37(1): 63-64. Krauss, Ellis S., and Robert J. Pekkanen. 2010. “The Rise and Fall of Japan’s Liberal Democratic Party.” The Journal of Asian Studies 69 (1): 5-15. Krook, Mona Lena. 2009. Quotas for Women in Politics: Gender and Candidate Selection

Reform Worldwide. New York: Oxford University Press. Krook, Mona Lena and Diana Z. O’Brien. 2012. “All the President’s Men? The Appointment of

Female Cabinet Ministers Worldwide.” The Journal of Politics 74(3): 840-855. LaFont, Suzanne. 2001. “One Step Forward, Two Steps Back: Women in the Post-Communist

States.” Communist and Post-Communist Studies 34(2):203-220. Levitz, Philip and Grigore Pop-Eleches. 2010. “Why No Backsliding? The European Union's

Impact on Democracy and Governance Before and After Accession.” Comparative Political Studies 43(4): 457-485.

Liddle, Joanna, and Sachiko Nakajima. 2000. Rising Suns, Rising Daughters: Gender Class and Power in Japan. New York: Zed Books.

Mair, Peter and Weeks, Liam. 2005. “The Party System.” In John Coakley and Michael Gallagher, eds.,Politics in the Republic of Ireland, 4 ed. New York: Routledge: 135-159.

Matland, Richard. 1998. “Women’s Representation in National Legislatures: Developed and Developing Countries.” Legislative Studies Quarterly 23(1): 109-125.

Matland, Richard E. 2002. “Enhancing Women’s Political Participation: Legislative Recruitment and Electoral Systems,” in Julie Ballington and Azza Karam, eds., Women in Parliament: Beyond Numbers. Stockholm: International IDEA: 1-13.

Matsui, Michiko. 1990. “Evolution of the Feminist Movement in Japan.” National Women’s Studies Association Journal 2(3), 453-449.

Murase, Miriam. 2006. Cooperation over Conflict: The Women’s Movement and the State in Postwar Japan. New York: Routledge, 26-44,

Ogai, Tokuko. 2001. “Japanese Women and Political Instituitons: Why Are Women Politically Underrepresented?” PS: Political Science and Politics 34(2): 207-210. Patessio, Mara. 2011. Women and Public Life in Early Meiji Japan: The Development of the

Feminist Movement. Ann Arbor: Center for Japanese Studies at The University of Michigan, 107-43.

Paxton, Pamela, and Melanie M. Hughes. 2007. Women, Politics, and Power: A Global Perspective. Thousand Oaks, CA: Pine Forge Press. Paxton, Pamela and Melanie M. Hughes. 2010. “Women as Presidents, Prime Ministers and

Government Ministers.” In K.P. O’Connor, ed., Gender and Women’s Leadership: A Reference Handbook. Thousand Oaks, CA: Sage.

Protsyk, Oleh. 2006. “Intra-Executive Competition between President and Prime Minister: Patterns of Institutional Conflict and Cooperation under Semi-Presidentialism.” Political Studies 54(2): 219-244.

Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000.

Page 54: Gender Volatility in Cabinets · recently, however, the lack of global time series data has made it difficult to explore systematic time trends for cabinets. The time series data

53

Democracy and Development. New York: Cambridge University Press. Reed, Steven R., Kenneth Mori McElwain, and Kay Shimizu, eds. 2009. Political Change in

Japan: Electoral Behavior, Party Realignment, and the Koizumi Reforms. Palo Alto: The Walter H. Shorenstein Asia-Pacific Research Center.

Reilly, Niamh. 2007. “Linking Local and Global Feminist Advocacy: Framing Women’s Rights as Human Rights in the Republic of Ireland.” Women’s Studies International Forum 30(1): 114-133.

Reynolds, Andrew 1999. “Women in the Legislatures and Executives of the World: Knocking on the Highest Glass Ceiling.” World Politics 51(4): 547-572.

Rosenbluth, Frances M., and Michael F. Thies. 2010. Japan Transformed: Political Change and Economic Restructuring. Princeton: Princeton University Press. Shugart, Matthew Soberg and John M. Carey. 1992. Presidents and Assemblies: Constitutional

Design and Electoral Dynamics. Cambridge: Cambridge University Press. Siavelis, Peter M. and Scott Morgenstern eds. 2008. Pathways to Power: Political Recruitment

and Candidate Selection in Latin America. State College, PA: Penn State University Press.

Smyth, Ailbhe. 1988. “The Contemporary Women’s Movement in the Republic of Ireland.” Women’s Studies International Forum 11(4): 331-341.

Stockemer, Daniel. 2009. “Women’s Parliamentary Representation: Are Women More Highly Represented in (Consolidated) Democracies than in Non-Democracies?” Contemporary Politics 15(4): 429-443.

Stockwin, J. A. A. 2008. Governing Japan: Divided Politics in a Resurgent Economy, 4 ed. Malden, MA: Blackwell Publishing. Svolik, Milan. 2008. “Authoritarian Reversals and Democratic Consolidation.” American

Political Science Review 102(2): 153-168. Viterna, Jocelyn, Kathleen M. Fallon, and Jason Beckfield. 2008. “How Development Matters: A

Research Note on the Relationship between Development, Democracy, and Women’s Political Representation.” International Journal of Comparative Sociology 49(6): 455-477.

Wangnerud, Lena. 2009. “Women in Parliaments: Descriptive and Substantive Representation.” Annual Review of Political Science 12: 51-69.

Welch, Susan and Donley T. Studlar. 1990. “Multi-Member Districts and the Representation of Women: Evidence from Britain and the United States.” The Journal of Politics 52(2): 391–412.

Williams, Z. 2011.”José Zapatero’s Feminista Agenda.” The Guardian. 31 March. http://www.guardian.co.uk/world/2011/apr/01/jose-zapatero-feminism-spain

Wright, Joseph and Abel Escriba-Folch. 2012. “Authoritarian Institutions and Regime Survival: Transitions to Democracy and Subsequent Autocracy.” British Journal of Political Science 42(2): 283-309.