ANRV377-PL12-04 ARI 13 April 2009 12:41

Women in Parliaments:Descriptive and SubstantiveRepresentationLena WangnerudDepartment of Political Science, University of Gothenburg, Gothenburg, Sweden;email: lena.wangnerud@pol.gu.se

Annu. Rev. Polit. Sci. 2009. 12:51–69

The Annual Review of Political Science is online atpolisci.annualreviews.org

This article’s doi:10.1146/annurev.polisci.11.053106.123839

Copyright c© 2009 by Annual Reviews.All rights reserved

1094-2939/09/0615-0051$20.00

Key Words

women’s interests, gender equality, established democracies, feministawareness

AbstractThis essay reviews two research programs. The first focuses on varia-tions in the number of women elected to national parliaments in theworld (descriptive representation), and the second focuses on effectsof women’s presence in parliament (substantive representation). Thetheory of the politics of presence (Phillips 1995) provides reasons forexpecting a link between descriptive and substantive representation.The safest position would be to say that results are “mixed” when itcomes to empirical support for the theory of the politics of presence.However, when a large number of studies covering a wide set of in-dicators on the importance of gender in the parliamentary process arepiled together, the picture that emerges shows that female politicianscontribute to strengthening the position of women’s interests.

51

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

INTRODUCTION

The number of women in parliaments is cur-rently increasing throughout the world. Tenyears ago the average proportion of womenin national parliaments was 11.7%; today it is18.3% (an increase of 6.6 percentage points).The trend is perhaps even clearer if, insteadof comparing world averages, we compare thenumber of countries where women make up>30% of members in the national parlia-ment. Ten years ago there was only one suchcountry—Sweden; now, as of September 30,2008, there are 22 (figures from the Inter-Parliamentary Union website http://www.ipu.org).

There is a growing body of literature thatseeks to explain the increases in the numbersof women elected. It is apparent that changesdo not happen automatically. Conscious acts byactors such as political parties, with the specificaim of getting more women elected, are an im-portant factor behind the increases. However,parties do not exist in a vacuum. The literaturereveals an interplay between parties and interestgroups such as women’s organizations, and alsobetween these kinds of actors and structures ofsociety. The type of electoral system matters forthe number of women elected, as does the typeof welfare state.

In research on women in parliaments, thereis a much-used distinction between descriptiveand substantive representation. This distinctionroughly corresponds with whether the focus ison the number of women elected or on theeffects of women’s presence in parliament. In1995 the influential book The Politics of Pres-ence was published (Phillips 1995). The theoryof the politics of presence suggests that femalepoliticians are best equipped to represent theinterests of women; thus, the theory predicts alink between descriptive and substantive rep-resentation. Phillips’s argument is built upondifferences between women and men in theireveryday lives, such as differences relating tochild-rearing, education and occupations, divi-sions of paid and unpaid labor, exposure to vi-olence and sexual harassment, and the fact that

female politicians, at least to some extent, sharethe experiences of other women. Few deny thatgender-related differences exist in contempo-rary societies; however, the connection to thepolitical sphere is disputed. Phillips herself usedthe expression a “shot in the dark” (1995, p. 83)in reference to expectations for female politi-cians to affect politics in any specific ways.Her doubt stemmed from her knowledge aboutrigidity in political institutions; parliaments donot change easily.

This essay on women in parliaments is orga-nized around the distinction between descrip-tive and substantive representation. I show thateven though the two research programs areinterconnected, they constitute distinct trackswith their own sets of theoretical reasoning andtools for doing empirical research. Research fo-cusing on descriptive representation representsa more mature research field. This is partly be-cause of the longer history of this tradition; asearly as the 1950s, Duverger (1955) highlightedthe role of electoral systems in determining thenumber of women elected. It is equally impor-tant to note that in research on descriptive rep-resentation there is a distinct, easily calculated,dependent variable: What is to be explainedis the numerical distribution of seats betweenwomen and men. Comparisons are made acrosscountries and across time. Comparisons are alsomade between subnational units, such as be-tween different parties or local legislatures.

Research on substantive representation isless mature. This is partly because there usedto be very few countries with any substantialnumber of women elected. There are, as al-ready hinted, additional complexities that, forexample, relate to the way parliaments func-tion as institutions. Some findings reveal that,once elected, female politicians meet certainobstacles. A fourth reason is that, in compar-ison to research on descriptive representation,the dependent variable is more diffuse or multi-faceted. It is not self-evident what an increasednumber of women in parliament will most af-fect: Internal working procedures? Policy out-comes? Trust in government? Or something

52 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

else? However, the body of literature in thisstrand of research is growing, and recent empir-ical results demonstrate that an increased pro-portion of women in parliament is more reliablethan a shot in the dark: Societies that elect largenumbers of women tend to be more gender-equal also in other respects than societies thatelect few women.

My review focuses on women in parliamentsin established democracies. However, especiallywhen it comes to research on descriptive rep-resentation, the perspective has to be widened.There is at present a global quota trend that hasresulted in rather surprising rankings amongcountries. The Nordic countries—as is widelyrecognized—hold a leading position, with theproportion of women in national parliamentsaveraging 41.4%. But among individual coun-tries, Rwanda is currently on top of the list(48.8%), Sweden is number two (47.0%), Cubanumber three (43.2%), and Finland numberfour (41.5%).

THEORIES ON WOMEN’SINTERESTS

The assumption that women have certain in-terests in common is a main thread in stud-ies on women in parliaments. However, in re-search focusing on descriptive representation,the theoretical reasoning behind this assump-tion is not well elaborated. In order to analyzenumeric differences between countries, or nu-meric changes over time, far-reaching defini-tions of women’s interests are not necessary. Itis widely acceptable to use the share of seats inparliament as an indicator of political inclusionin society for a category such as women. In thisstrand of research, the theoretical focus is in-stead directed toward developing explanationsfor variations.

In studies focusing on substantive represen-tation, it is, however, necessary to specify con-cepts such as “women’s interests” and “genderequality.” A core idea in this strand of research isthat there are certain interests and concerns thatarise from women’s experiences and that thesewill be inadequately addressed in a politics that

is dominated by men. However, the concept ofwomen’s interests is contested. Contemporarydebates concern features of elitism in genderresearch—that is, a tendency to ascribe inter-ests to women in a top-down fashion—and alsofeatures of essentialism: the tendency to viewwomen and men as fixed, rather than change-able, categories. Debates also concern how gen-der is related to categories such as ethnicity, age,and class (Dietz 2003).1

One way to handle such controversies isto let politically active women themselves de-fine women’s interests or what they perceiveas gender equality; this strand of research re-lies on what are labeled “subjectively definedinterests” (e.g., Celis 2006). Other researcherselaborate theoretically founded definitions thatare sensitive to diversity among women but alsostate some common ground (e.g., Lovenduski& Norris 2003, Wangnerud 2000a). In prac-tice, these different approaches often producesimilar lists of women’s interests. Phillips’s rea-soning in The Politics of Presence is an example ofmainstream argumentation:

Women have distinct interests in relation tochild-bearing (for any foreseeable future, anexclusively female affair); and as society is cur-rently constituted they also have particular in-terests arising from their exposure to sexualharassment and violence, their unequal po-sition in the division of paid and unpaid la-bor and their exclusion from most arenas ofeconomic or political power. (Phillips 1995,pp. 67–68)

When Phillips concretizes what women cangain from increased political inclusion, shestresses context; women’s interests are con-nected to how societies are currently consti-tuted. If we look at contemporary societies, wesee noteworthy differences between women and

1The concept of interests is not limited to gender research. Itis “ubiquitous” (Pitkin 1967, p. 156) in debates on represen-tation. To differentiate interests is a matter of concretizingthat which various groups can expect to gain through politicalinclusion.

www.annualreviews.org • Women in Parliaments 53

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

men in their everyday life situations. It is there-fore interesting to ask about the extent to whichan increase in the number of women elected af-fects how societies function. The contextual ap-proach implies that concepts such as women’sinterests and gender equality are anchored intime and space; this means that more exact def-initions have to be worked out in relation to theactual parliament studied. One problem here,though, is that even the most carefully contex-tually anchored definition will necessarily endup a bit simplified. However, when I review re-cent empirical research on substantive repre-sentation of women I conclude that this is a riskworth taking. Gender serves as a lens that makesimportant issues in the field of representationvisible: Whom do elected politicians represent?What is at stake in the parliamentary process?What do we know about the interplay betweenparliaments and the everyday lives of citizens?Gender research adds new fuel to all these clas-sic debates.

In research on substantive representationof women, an aspect of politicization is intro-duced. It is commonly argued that societies willnot achieve equality between women and menby simply disregarding gender-related differ-ences (Phillips 2007, p. 127). This is not alwaysspelled out clearly in the literature; however, Ithink most authors would agree that in gender-equal democracies, women and men are equallyable to choose between alternatives that addresstheir specific concerns. The theory of the pol-itics of presence stipulates that equal rights toa vote are not strong enough to guarantee this;there must also be equality among those electedto office.

RESEARCH ON DESCRIPTIVEREPRESENTATION: NUMBEROF WOMEN IN PARLIAMENTS

Owing to a recent worldwide quota trend, con-ventional wisdom regarding explanations forvariations in the number of women elected ispartly outdated. However, it is still valid to re-view more established explanatory factors anddiscuss their strengths. In this section, I first

review research on gender and parliamentaryrecruitment in Western democracies and thenwiden the perspective. The discussion is re-stricted to factors connected to national par-liaments’ lower or single houses.

Norris (1993, p. 311) has worked out amodel of parliamentary recruitment in Westerndemocracies emphasizing the political system,the party context, and supply and demand fac-tors in the recruitment process. Factors in-cluded under the label “political system” arethe electoral system, party system and legisla-tive competition. Factors included under the la-bel “party context” are party ideology and partyorganization. The picture of the recruitmentprocess includes social background as a deter-minant for required resources to become anelected representative as well as for motivation.Resources and motivation are then describedas decisive factors for who gets into the poolof the eligible. Further obstacles that must beovercome before one gets elected are the judg-ments of gatekeepers and voters. The modelalso indicates that the outcome of the electionhas a feedback effect on motivation; if only afew women are elected, or if women are electedonly occasionally, this might enforce the ideathat politics is “a man’s game.”

One of the most stable results in empiricalresearch is that the election of women is fa-vored by electoral systems with party lists, pro-portional representation (PR), and large districtmagnitudes. The conventional wisdom used tobe that these systems are less competitive thanmajority systems based on single-member dis-tricts. In a single-member system, a woman hasto be the number-one choice for her party inorder to take part in the race. In a PR systemwith large district magnitudes, a woman can beplaced further down on the party list and still beelected (Matland & Brown 1992, Norris 1996,Rule 1987). However, this wisdom also main-tains that PR tends to favor party systems with alarge number of parties and also means greaterpossibilities for new parties to enter the parlia-mentary arena. In this sense a PR system is verycompetitive. Empirical studies show that onceone party picks up the issue of gender equality,

54 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

and makes the increased election of women anissue, other parties within the same system tendto follow suit. This logic can be explained bypolicy diffusion, a wish to pick up new ideas,and strategic considerations; parties are strivingfor power, and gender equality is one among anumber of weapons that can be used in elec-toral competition (Kittilson 2006, Lovenduski& Norris 1993).

There is no doubt that the political systemis relevant in cross-country research. However,cross-country studies tend to miss variations be-tween parties within a single system. Variationsin the proportion of women to men are evengreater across parties than across nations. Oneof the earliest findings in the field of party con-text was that leftist ideology is a strong pre-dictor for a high number of women elected.However, this pattern is weaker than it usedto be. Substantial increases are found in mostparty families in Western democracies, eventhough religious and ultraright parties still havefew women elected (Kittilson 2006). Kittilson(2006, p. 48) has distinguished between “oldleft” and “new left” political ideologies, andher analysis shows that new left ideology isa decisive factor in cross-party, cross-countrycomparative research. Parties that value envi-ronment over economic growth and are pro-permissive in social policy tend to have morewomen elected. In addition to party ideology,party organization matters: A centralized orga-nization is favorable for women, but even moreimportant is that the party organization has tieswith organizations outside the party, becausethese provide more points of access for women(Kittilson 2006, pp. 48–49).

To be elected, women must first be willingto stand as candidates. However, recent studiesshow that the idea of a smooth process in whichhigher numbers of women in the pool of the eli-gible automatically spill over to higher numbersamong the elected representatives is too simple.In Western democracies, the number of womenelected, in many cases, has not grown incremen-tally but rather in “punctuated and sometimesdramatic increases” (Kittilson 2006, p. 10). Thisfinding puts internal processes within parties,

the main gatekeeper in the recruitment process,at the center of the analysis.

Recent studies point to conscious acts byparty leadership, such as the implementationof gender quotas or other gender-specific mea-sures, as important determinants of the numberof women elected (Freidenvall 2006, Kittilson2006, Studlar & McAllister 2002). During pastdecades, the number of women elected has alsoincreased in parties that have never adoptedquotas; however, the average number of womenelected is significantly higher in parties thathave adopted quotas. Kittilson (2006) compares71 parties in Western Europe between 1975 and1995. In 1975 few parties had adopted quotas,and the average number of women elected wason the same level, about 10%, in each of theparty groupings in her study. In 1995 the aver-age number of women elected had increased to35% in parties with quotas (21 parties in thesample) and to 25% in parties without quo-tas (50 parties in the sample). She divides par-ties with quotas into two subgroups: those thatadopted quotas between 1975 and 1985, andthose that adopted quotas between 1986 and1991. Kittilson (2006, p. 64) demonstrates thataverages within quota parties, in each sepa-rate subset, jumped above averages within non-quota parties shortly after quotas were adopted.Parties with quotas also remain on a high aver-age level.

The literature on causes behind the elec-tion of large numbers of women reveals a shift,at least in Western democracies, from system-oriented toward strategy-oriented explanatorymodels (Wangnerud 2000b). Strategy-orientedexplanatory models dig into processes endoge-nous to parties, whereas system-oriented ex-planatory models explore exogenous processes.This shift does not mean that the frameworkfor parliamentary recruitment worked out byNorris has become obsolete; it still serves as agood overview and spells out important factors.However, other layers or steps in the processhave to be developed more fully. Timing andframing are two examples of concepts used instrategy-oriented research to capture the mech-anisms at work. The strategy-oriented strand

www.annualreviews.org • Women in Parliaments 55

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

of research also points to the importance of ac-tions taken by women themselves, often in ajoint venture between women’s movements andwomen at higher ranks within the party struc-tures. Timing and framing are strategies to in-crease support within party machineries; if thetime is “right” and the packaging “appetizing,”quests for increased gender equality will have afairly good chance of gaining support (Kittilson2006, Lovenduski & Norris 1993).

Cultural Explanationsand Socioeconomics

From the start of cross-country research onwomen in parliaments, regional differenceshave been a recurrent finding. The high per-centage within the Nordic region (41.4%) isexceptional, and perhaps even more remarkableis that within this region the number of womenelected has been high for quite a long time. Inthe 1980s the expression “Norden—the passionfor equality” was coined (Graubard 1986). Thisslogan alludes to values deeply embedded insociety.

Gender culture can be defined as societalideals, meanings, and values that have genderconnotations (Pfau-Effinger 1998). In the bookRising Tide: Gender Equality and Cultural Changearound the World, Inglehart & Norris (2003)emphasize the importance of a gender-equalityculture in which women have opportunities forupward mobility. The opposite of a gender-equality culture is one where traditional gen-der values prevail. Inglehart & Norris constructa gender-equality scale from measurements ofcitizens’ attitudes regarding women as politicalleaders, women’s professional and educationalrights, and women’s traditional mother role.The empirical findings show that the gender-equality scale correlates with the number ofwomen in parliament (Inglehart & Norris 2003,p. 138).2

2The correlation is strong (0.57, level of significance .01),yet there are outliers to the general pattern; some establisheddemocracies like Australia and the United States displaymore egalitarian attitudes than might be expected given the

Inglehart & Norris are not the first to em-phasize culture, and the major contributionof their book is the vast amount of empiricaldata they provide, covering almost all parts ofthe world. However, the cultural explanationhas been criticized for being unable to cap-ture short-term change. It has also been crit-icized for being almost a tautology (Sainsbury1993, Rosenbluth et al. 2006). Nonetheless, re-gional differences are a persistent feature ofworldwide comparisons, and additional studiessubstantiate the view that gender-related valuesare important in this field (Paxton & Kunovich2003, Studlar & McAllister 2002, Teigen &Wangnerud 2009).

A pattern in the literature is that researchersdoing worldwide comparisons on the numberof women elected emphasize different factorsfrom those emphasized by researchers doingmore geographically restricted analyses. Thetime perspective used is also important in de-termining the kinds of factors that stand outas decisive. Wide’s (2006) analysis includes 74to 179 countries during the postwar period1950–2005; numbers of countries vary accord-ing to how many states are independent andalso have parliaments. Wide shows that untilthe fall of the Berlin Wall in 1989, the year ofthe introduction of female suffrage and the ap-pearance of communist regimes are the mostdecisive factors. The communist regimes inEastern and central Europe practiced a sys-tem of gender quotas. In the period after 1990,proportional electoral system is the most deci-sive factor and, in addition to the use of gen-der quotas, has a significant effect. Wide (2006)also shows that during the whole period from1950 to the present, Protestantism and highlevels of economic development correlate posi-tively with a high number of women elected (seealso Studlar & McAllister 2002). Other large-nstudies show that socioeconomic factors, suchas women’s share in professional occupations,

proportion of women elected to parliament. Inglehart &Norris (2003, p. 139) conclude that in those countries “publicopinion seems to run ahead of the opportunities that womanhave when pursuing public office.”

56 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

have an impact (Kenworthy & Malami 1999,Matland 1998, Salmond 2006).

The Recent PostcommunismQuota Trend

In the book Women, Quotas and Politics,Dahlerup (2006a) and colleagues analyze therecent postcommunism quota trend. Genderquotas are generally understood as formal-ized measures with the specific aim of increas-ing the number of women elected. There are40 countries where gender quotas in electionsto national parliaments have been implementedby means of constitutional amendment or bychanging the electoral laws; these are legal quo-tas. In another 50 countries, major politicalparties have set out quota provisions in theirown statutes; these are party quotas (Dahlerup2006a; for current research debates on genderquotas, see theme articles collected under thetitle Gender Quotas I & II in Politics & Gender2005 and 2006 respectively).

The recent quota trend can be traced toNorway at the beginning of the 1970s, when theSocialist Left Party implemented gender quotasregarding seats in internal party settings, such asthe Party Board, as well as external party candi-date lists. However, many observers point outthat it was the United Nations Fourth WorldConference on Women in Beijing, 1995, thatsparked changes. The use of gender quotas isbecoming especially frequent in Latin Americaand sub-Saharan Africa. Dahlerup (2006a, p. 4)suggests an international contagion effect wasimportant for the spread; for some countries,the implementation of quotas reflects a wishto appear “modern” in the international com-munity. Dahlerup reflects that state-driven po-litical inclusion of women might foremost besymbolic.

The present quota trend causes a specialkind of historical leap in many economicallyless developed countries. As mentioned above,Kittilson (2006) states that dramatic increasesare also a part of the story in Western democra-cies. But in other parts of the world, there is aneven stronger divergence between changes in

the number of women elected and changes inthe status of women in society more generally.

Dahlerup (2006a) and colleagues distin-guish between fast-track and incremental mod-els regarding the number of women elected;however, there are no clear boundaries here.Whether changes are seen as fast or not de-pends on the reference points used. However,if the development in Sweden is contrasted withthe development in Rwanda, the differences inthe models are striking. By the beginning of the1970s, parties in Sweden had started to imple-ment measures, often referred to as soft quo-tas, in order to increase the number of womenelected. During the 1970s Sweden crossed thethreshold of 20% women in parliament; theproportion climbed past 30% during the 1980sand 40% during the 1990s. This step-by-stepdevelopment, spanning almost four decades,lies behind the current figure of 47.0% womenin the Swedish parliament (Bergqvist et al.2000, Freidenvall 2006). Rwanda’s situation ismuch different. Whereas Sweden’s twentieth-century history is characterized by political sta-bility, economic growth, and peace, Rwandais one of the poorest countries in the world,and its modern history contains disastrous wars.Gender quotas were implemented in Rwandaas a part of the reconciliation process after thegenocide in 1994. In 1994 women made up17.1% of the national parliament in Rwanda.After the election in 2003, the number was48.8%. The number of women tripled in lessthan ten years (Devlin & Elgie 2008).

Corruption, Welfare State,Labor Market Characteristics,and Voter Preferences

Before ending this section on descriptive repre-sentation, I draw attention to some additionalfactors discussed in the literature on womenin parliaments: corruption, the type of welfarestate, labor market characteristics, and voterpreferences. These factors are shedding newlight on the processes at work; however, theyare not yet firmly included in the researchcanon.

www.annualreviews.org • Women in Parliaments 57

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Studies on corruption, such as those initi-ated by researchers at the World Bank, find ev-idence of a relationship between the number ofwomen in parliament and the level of corrup-tion: The higher the number of women in thenational parliament of a country the lower thelevel of corruption (Dollar et al. 2001). How-ever, the causal direction of the relationship isnot clear, and Sung (2003, p. 718) observes that“[g]ender equality and government account-ability are both great achievements of mod-ern liberal democracy.” Sung confirms, though,that the degree of “fairness” or “cleanness” in asystem is of importance for women’s ability toachieve politically important positions.

Rosenbluth et al. (2006) use Nordic/Scandinavian exceptionalism as the startingpoint of an analysis intended to explain themechanisms by which the expansion of the wel-fare state facilitates women’s entry into politics.The Nordic region is characterized not onlyby the high proportion of women elected toits parliaments but also by its encompassing“cradle-to-grave” welfare states. The key linkbetween female legislative representation andthe Nordic type of welfare state is presented inthree steps: Welfare state policies free womento enter the paid workforce; they provide publicsector jobs that, to a disproportionate degree,employ women; and hence the political inter-ests of working women are changed enough tocreate an ideological gender gap (see Hernes1987).

The results reported by Rosenbluth et al.(2006) show that increases in government (non-military) expenditure are consistently associ-ated with increases in the number of womenelected, and the results hold when controlledfor factors such as left ideology and female la-bor force participation. However, the partiesare still gatekeepers in the recruitment process.Increasing the number of female candidates isdescribed by Rosenbluth and colleagues as a“fruitful” step for a party to take in order to“exploit” the gender gap created by a new set ofpreferences.

The case of the United States presents aparadox. The number of women elected to the

U.S. Congress is lower, 16.8% in 2008, thanwould be predicted from female labor forceparticipation and egalitarian attitudes amongcitizens. Iversen & Rosenbluth (2008) suggestthat the political arena can be analyzed as ajob-market arena. In candidate-centered po-litical systems, such as the U.S. system, se-niority and uninterrupted careers matter morethan in party-centered political systems, andconstituency services are highly emphasized.Women aspiring to a parliamentary position arehit harder by the need to take time off for childcare and other family duties.

Some research on gender differences amongvoters relates to research on women in parlia-ments. One important finding from time-seriesapproaches is that gender differences amongthe citizenry tend to be pervasive; they re-flect societal conditions and exist outside thecontext of specific elections (Box-Steffensmeieret al. 2004). In addition, the dynamics of aspecific election—whether high-profile womentake part and/or whether women’s issues arehigh on the agenda—affect these differencesor gender gaps. Ondercin & Bernstein (2007,p. 50) conclude, “The voice of women in theelectorate is heard more loudly when a womanarticulates the views on which women and mendiffer.” However, the specific issues that gen-erate gender differences in each election vary(Cutler 2002, Dolan 2005, Koch 2002). The re-search instrument plays an important role; thekinds of gender-related patterns that arise re-late to the kinds of issues that are included, forexample, in questionnaires.

There is some evidence that female votersare more supportive of female politicians thanmale voters are (Banducci & Karp 2000). How-ever, some findings indicate that female vot-ers tend to overestimate the actual number ofwomen in parliamentary positions to a largerextent than male voters do. If our knowledgewere more accurate, the issue of gender equalityin parliament would perhaps be more stronglyprioritized (Sanbonmatsu 2003).

The literature on descriptive representationof women brings forward a broad spectrum ofexplanatory factors. I perceive the challenge

58 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Table 1 Explanatory factors brought forward in research on variations in the number of women electedto national parliaments (descriptive representation)

Macro-level variables Meso-level variables Micro-level variablesyear of female suffrage party ideology voter preferencesreligion party organization motivation among women to be candidatesgender-equality culture party gender quotassystem cleanness women’s movementelectoral system timing and framingwelfare state system dynamics in specific electionslegal gender quotas contagion effects across partiesgovernment (nonmilitary) expendituresocioeconomicscontagion effects across countries

for future research to be the building of fine-tuned models that specify relationships. Eventhough research in this field is mature and usessophisticated methods for data analysis, manystudies test the impact of one “favorite” factorwhile controlling for some “standard” factors.Macro-level variables still dominate research onwomen’s descriptive representation. However,a promising trend is found in research (e.g.,Kittilson 2006) that scrutinizes the interplaybetween structures of society and actors suchas political parties and women’s movements.Table 1 facilitates an overview of explanatoryfactors brought forward in research on descrip-tive representation.

RESEARCH ON SUBSTANTIVEREPRESENTATION: EFFECTS OFWOMEN IN PARLIAMENTS

What do women do in parliaments? In mostWestern democracies, it is possible to findprominent female politicians in areas such asforeign affairs and finance, as well as in areassuch as education or family policy. However,the core issue in research on substantive repre-sentation does not concern “what women do inparliaments” but, more specifically, the extentto which the number of women elected affectswomen’s interests. Phillips (1995, p. 47) arguesthat gender equality among those elected to of-fice is desirable because of the changes it can

bring about: “It is representation . . . with a pur-pose, it aims to subvert or add or transform.”

In the introduction to this essay, I statedthat research on substantive representation ofwomen is a less mature field than researchon descriptive representation. One indicator ofthis is the lack of agreement in the literatureon what effects to expect when the number ofwomen in parliament increases. In fact, much ofthe current research debate on substantive rep-resentation of women concerns hindrances tofemale politicians. The point is made that sub-stantial change—whatever that means—cannotbe taken for granted just because a group, suchas women, is taking part in decision making to alarger extent than before. The link between de-scriptive and substantive representation that issuggested by the theory of the politics of pres-ence is probabilistic rather than deterministic(Dodson 2006).

The relationship between descriptive andsubstantive representation is hard to capture,but the problems in this field of research shouldnot be overemphasized. The review in this sec-tion shows that a reasonable set of indicators isbeing used in empirical research and that theseindicators can serve as a base for further de-velopments. Before I get to research on effectsof the increased number of women elected, Ireview research on hindrances to female politi-cians. This research is concerned with precon-ditions for change.

www.annualreviews.org • Women in Parliaments 59

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Preconditions for ChangeThe concept of “critical mass” is intensely de-bated. Some scholars seek to identify a thresh-old number or a “tipping point” at which theimpact of women’s presence in parliaments be-comes apparent; a figure of ∼30% is often men-tioned. However, others criticize the concept ofcritical mass as being too mechanical and imply-ing immediate change at a certain level. Theyfocus instead on “critical acts” (Dahlerup 1988)to explore two questions: Who is pushing forchange consistent with women’s interests, andwhat kinds of strategies are useful (Dahlerup2006b; Childs & Krook 2006)? Still others (e.g.,Grey 2006) suggest that different thresholdshave to be recognized in studies on women inparliaments; for example, attaining a propor-tion of 15% may allow female politicians tochange the political agenda, but 40% is neededfor women-friendly policies to be introduced.

Hindrances to female politicians, such ashostile reactions to women, working conditionsincompatible with family responsibilities, andthe existence of male-dominated networks aresuggested in the literature. Lovenduski (2005a,p. 48) argues that the most difficult obstacleis the deeply embedded culture of masculinityin political institutions. She points out genderbiases in personnel, in policy, and in culturesof political organizations (Lovenduski 2005a,p. 52).

The question of how the presence of womenaffects behavior and culture within politicalinstitutions is multi-layered. The question isnot just about whether women behave differ-ently, or whether they meet certain obstacles,or whether, beyond a certain threshold ofnumbers, they are able to make an impact.The question is also whether their presencehas an impact on the behavior of men, eitherreinforcing gender differences or modifyingthem. In addition, it is important to take intoaccount some less gender-specific factors in theanalyses of impact. Grey (2006) points to theimportance of the institutional positions of thefemale politicians in question, the time womenhave spent in office, and their own as well as

their political party’s ideology. Beckwith (2007)summarizes much of the discussion by statingthat both numbers and newness must be takeninto account when analyzing gender; beingnew in parliament is widely recognized as afactor that diminishes possibilities for impact.

The importance of newness is supported byempirical results from Jeydel & Taylor (2003),who show that when factors such as senior-ity and institutional position are taken into ac-count, there is no real demonstrable differencebetween the effectiveness of women and men inthe U.S. House of Representatives. Two mea-sures of effectiveness are used in their study:percentage of bills sponsored by a member ofCongress that were passed into law and the dis-tribution among congressional districts of fed-eral money to implement domestic policy.

However, the suggestion that impact in par-liaments is related not only to levels of senior-ity but also to gender understood in a “purer”sense also coincides with results from empiricalresearch. Haavio-Mannila et al. (1983) distin-guish two kinds of divisions between womenand men: those related to formal power (hi-erarchical gender structures) and those relatedto policy areas (functional gender structures).Research on functional structures has focusedon parliamentary committee assignments, andthe patterns displayed with respect to feminin-ity and masculinity are puzzling.

Thomas (1994) is a pioneer of empirical re-search on gender and committee assignments.In an analysis on state legislators in the UnitedStates, she follows developments over time: Inthe 1970s, women representatives were concen-trated in a very narrow set of committees, mostoften education committees; however, in the1980s, women were found in all kinds of com-mittees. However, the proportion of womenand men was not equal on all types of com-mittees. A 1988 survey showed that womenwere significantly more likely than men tobe assigned to health and welfare committees;women were also less likely than men to siton committees dealing with business and pri-vate economic concerns (Thomas 1994, p. 66).

60 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Thomas also investigates the extent to whichcommittee assignments reflect priorities amongmale and female politicians. Her conclusionfrom the 1988 survey was that gender pat-terns resulted from legislators’ choices ratherthan coercion or discrimination (Thomas 1994,p. 67).

In studies on committee assignments in theSwedish parliament, the time perspective iswider. Wangnerud (1998) focuses on stand-ing committees during the period 1970–1996.The data reveal that the division between maleand female policy areas was especially clear-cut in the 1980s and early 1990s. However,the Swedish election of 1994 was followed by anotable change; previous gender patterns, verymuch in line with those reported by Thomas(1994) for U.S. state legislators, almost dis-appeared. In the present Swedish parliament,there is only one standing committee with<40% women, and that is the committee onsocial insurance, a formerly heavily female-dominated area (L. Wangnerud, unpublishedmaterial).

Several conclusions can be drawn from re-search on gender and committee assignments.The first is that patterns of functional divisionsbetween women and men in parliaments can-not be satisfactorily explained by factors suchas the number of years politicians have spent inparliament, their party affiliation, or their age.The overall picture is that functional divisionsstem from men’s and women’s different pref-erences for committees. However, there is alsoempirical support for the existence of stereotyp-ing processes; women displaying typical “male”preferences met special obstacles, for example,in the Swedish parliament during earlier timeperiods. Another important finding from time-series approaches is that the magnitude as wellas the shape of the functional division variesover time. It is also important to state that thereare no obvious correlations between functionaldivisions and the status of different policy areas(Wangnerud 1998; for a different view on statussee Heath et al. 2005).

The literature provides no clear-cut answerto the question of what obstacles female politi-

cians meet. The most important conclusionto be drawn from my review of research onhindrances for female politicians is that nostrong empirical evidence supports the idea thatwomen are merely symbolic representatives. Itwould be far-fetched to say that women’s roomto maneuver in parliaments in Western democ-racies is severely undermined by their gender.

Indicators of Attitudes, Priorities,and Policy Promotion

The point of departure for the theory of the pol-itics of presence is sociological. Female politi-cians are expected to be better equipped torepresent the interests of female voters be-cause they, at least to some extent, share thesame experiences. There are plenty of counter-hypotheses to this expectation—for example,that ideology is what matters in politics, that so-cial characteristics such as class or ethnicity aremore decisive than gender, and that parliamen-tary institutions influence politicians more thanpoliticians are able to influence them (meaningthat women entering parliament become justlike the male politicians who preceded them).

It goes without saying that parliaments arecomplex institutions and that it is a method-ological challenge to empirically test the the-ory of the politics of presence. One suggestionis that studies in this field ought to be longitu-dinal in design; we should follow what happens“from the start” when women are few, up to thepoint where women are present in large num-bers (Beckwith 2007). Longitudinal designs ofthis kind are hard to conduct. An alternative isto use a wide range of indicators in cross-sectionanalysis and include control variables in orderto isolate effects of gender.

Thomas (1994) does cross-section researchthat includes more than one time-point and usesa wide set of indicators. In her book How WomenLegislate, she distinguishes between legislativeprocedures and legislative products. Legislativeprocedures include activities such as makingspeeches, working with colleagues, and bar-gaining with lobbyists. Legislative products in-clude voting records, issue attitudes, and policy

www.annualreviews.org • Women in Parliaments 61

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

priorities. Thomas’s findings show a closinggap between women and men concerningprocedures; however, when it comes to prod-ucts, the gap does not appear to be closing.“Today,” Thomas (1994, p. 7) concludes,“women legislators embrace priorities dealingwith issues of women, and children and thefamily. Men do not share this priority list.”

Recent empirical research demonstrates anambition to further develop measurements forlegislative products. Some choose to structuretheir analysis around dimensions such as pol-icy style, agenda, and outcomes (Squires &Wickham-Jones 2001), and others choose di-mensions such as legislative voting, parliamen-tary roles, and ideological values (Lovenduski& Norris 2003). I have chosen to structure thisreview around attitudes, priorities, and policypromotion because these dimensions capturethe bulk of recent empirical research.

Gender differences in politicians’ attitudeshave been studied rather frequently. Thereis an agreement in research that gender hasan impact; what varies between studies is thestrength of the impact (Esaiasson & Holmberg1996, Lovenduski & Norris 2003, McAllister &Studlar 1992, Narud & Valen 2000, Norris &Lovenduski 1995). Even though the magnitudeis disputed, the direction of the gender gap inattitudes is fairly clear: Women in parliamentstend to be more leftist than men, and they tendto be more favorable toward new policies, suchas those concerned with environmental protec-tion. Differences also appear on issues that canbe defined as women’s interests, such as socialpolicy (women support more permissive poli-cies); pornography (women are more skeptical);and affirmative actions such as introducing gen-der quotas (women are more in favor). Sugges-tions have been made that gender differencesappear on issues that are not yet central to par-ties (Heidar & Pedersen 2006).

Research on attitudes highlights the impor-tance of party affiliation. Some attitudes, suchas those regarding the pornography issue, tendto split women and men regardless of theirparty, whereas other issues split women andmen in some parties but not in others. In a

study on gender differences in U.S. state legisla-tors’ policy preferences, Poggione (2004) foundthat women expressed significantly more liberalwelfare policy preferences than men. The dif-ference was most profound among Republicanlegislators.

A variation on measuring attitudes is to re-search how parliamentarians define their task.Female politicians, to a larger extent than theirmale colleagues, view the “representation ofwomen’s interests” (which has been includedin questionnaires) as part of their duty. In astudy of the five Nordic countries, Esaiasson(2000, p. 64) analyzes behavioral consequencesof such a task definition, and he concludes that“self-defined champions” of women’s interestsare more inclined than others to contact cab-inet ministers on behalf of women. Dodson(2006) reports similar results regarding the U.S.Congress.

Research on attitudes examines what so-lutions are favored once an issue is on thepolitical agenda. In contrast, research on pri-orities focuses on an earlier step, asking whichissues get onto the political agenda in the firstplace. Empirical research shows that femalemembers of parliament tend to prioritize issuesthat are also prioritized by female voters (Diaz2005, Reignold 2000, Skjeie 1992, Swers 1998,Thomas 1994, Wangnerud 2000a).

Wangnerud’s (2006) study of gender differ-ences in priorities among Swedish parliamen-tarians covers a period of almost 20 years. Thefocus is on the number of male and femalepoliticians who mention social policy, familypolicy, care for the elderly, or health care as acampaign issue or an area of personal interest.The items included in the study can be seen asa broad way of conceptualizing priorities classi-fied as women’s interests, since policies in theseareas have a special bearing on the everydaylife of women. The empirical analysis showsthat in 1985, 75% of female members of par-liament addressed issues of social policy, familypolicy, care for the elderly, or health care in theirelection campaigns. The corresponding figureamong male members of parliament was 44%(a difference of 31 percentage points). In that

62 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

same year, 52% of female members of parlia-ment stated that issues of social policy, familypolicy, care for the elderly, or health care wereareas of personal interest, compared with 11%of male members of parliament (a difference of41 percentage points). These noteworthy gen-der differences hold when the results are con-trolled for party affiliation and age.

Another important result from this research(Wangnerud 2006) is that, although genderdifferences are found on the two subsequentsurvey occasions, 1994 and 2002, the gap hasnarrowed over time. An additional significantresult in the Swedish study is that genderdifferences were much greater in personalagendas (areas of personal interest) than incampaign agendas. In Swedish politics, theelection campaign is centrally controlled toa rather great extent by the parties. Whenanalyzing effects of gender, it is importantto note that different political arenas can beexposed to different levels of outside influence.

The results presented here should not be in-terpreted to mean all women in parliament nec-essarily promote the same kinds of solutionsto social problems. In a study of the Norwe-gian parliament, Skjeie (1992) analyzes how ef-fects of gender are filtered through party ide-ology. Her results show that more women thanmen give priority to issues of “care-and-careerpolitics”—how to successfully combine familylife and working life; however, women fromright-wing parties tend to support private so-lutions, whereas women from left-wing partiestend to support state intervention.

Priorities can also be measured in termsof the parliamentarian’s contacts. I referred tothe study of Esaiasson (2000), which showedthat more female than male members of par-liament in the Nordic countries have frequentcontacts with cabinet ministers on behalf ofwomen. In addition, a study of the Nordicparliaments found that female politicians havemore frequent contacts with women’s organi-zations outside parliament. Women in Nordicparliaments also cooperate across party lines inorder to influence parliamentary decisions re-lated to gender equality to a greater extent than

their male colleagues. It is obvious in studiesof the Nordic countries that women in parlia-ments give the issue of gender equality higherpriority than their male colleagues, even thoughfar from all women are “champions” of equality(Wangnerud 2000b; cf. Bergqvist et al. 2000).

I have made some references to empirical re-search on political behavior; however, most ofthe research I have reviewed so far has relied onparliamentarians’ responses to questionnaires.Vega & Firestone (1995) have examined leg-islative voting behavior from 1981 to 1992 inthe U.S. Congress, and their results confirmthe findings of questionnaire-based research.They conclude that “congressional women dis-play distinctive legislative behavior that por-tends a greater representation of women andwomen’s issues” (Vega & Firestone 1995,p. 213). This finding is in line with Celis’s (2006)analysis of speeches from the budget debatesof the Belgian Lower House during the pe-riod 1900–1979. Female members of the Bel-gian Parliament were found to be women’s most“fervent” representatives (Celis 2006, p. 85).

Grey (2002) has conducted a study onchanges in New Zealand’s parental leave poli-cies. Changes were made in 1975 and 1999, andduring this time period, the number of womenin the parliament of New Zealand grew from<5% to almost 30%. Grey’s analysis of parlia-mentary debates preceding the changes showsa somewhat different result than, for example,Celis’s study of the Belgian Lower House: Themost obvious division in Grey’s study was alongparty lines and not along gender lines. Grey(2002) does, however, point to some importantchanges that have occurred in New Zealandparallel to the increase in the number of womenelected. As one example, leave provisions forboth parents, rather than mothers alone, havebeen accepted.

Although there are studies on policy pro-motion, the closer one gets to outcomes in cit-izens’ everyday lives, the fewer empirical find-ings there are to report. In a statistical analysisof child-care coverage in Norwegian munici-palities for 1975, 1979, 1983, 1987, and 1991,Bratton & Ray (2002) demonstrate that the

www.annualreviews.org • Women in Parliaments 63

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

number of women elected has influenced publicpolicy outputs (increased child-care coverage),but the effect of women representatives is notconstant; it is most evident in a period of policyinnovation. Bratton & Ray (2002) also point outthat an important precondition for the transla-tion of descriptive representation to policy out-comes is the existence of gender differences inthe mass public and the presence of women inthe executive.

The lack of research on outcomes in ev-eryday life for citizens is a problem. Especiallyunfortunate is the lack of good cross-countrycomparative research. However, a study bySchwindt-Bayer & Mishler (2005), using datafrom 31 democracies, serves as a promisingexample. The starting point is an integratedmodel of women’s representation that takes intoaccount both descriptive and substantive repre-sentation. Indicators of substantive representa-tion are weeks of maternity leave, indexes onwomen’s political and social equality, and mari-tal equality in law. The main conclusion is thatincreased descriptive representation increaseslegislatures’ responsiveness to women’s policyconcerns and also enhances perceptions of le-gitimacy among the electorate, but the authorsperceive the effects of substantive representa-tion to be smaller than anticipated in theory.

Different international organizations havecreated gender-equality scales, capturing as-pects such as female life expectancy, years offormal schooling for females, ratio of female-to-male earned income, and maternity leavebenefits. These indexes have to be evalu-ated further; however, worldwide compar-isons reveal some interesting parallels to thenumber of women elected. In 2007, Nordiccountries topped lists of the most gender-equal countries compiled by the United Na-tions (Gender-related Development Index), theWorld Economic Forum (Gender Gap In-dex), the International Save the Children Al-liance (Mothers’ Index), and the Social Watch(Gender Equity Index). Nordic countries alsotop lists that rank countries by the numberof women in parliament (Inter-ParliamentaryUnion website http://www.ipu.org).

Outcomes can also be understood in morenarrowly defined political terms. Atkeson& Carrillo (2007), having pooled AmericanNational Election Studies from 1988 to 1998,confirm that higher numbers of women electedpromote higher values of external efficacy forfemale citizens. The results hold when con-trolled for factors such as political participation,strength of partisanship, various electoral char-acteristics, and state political culture. Politicalefficacy is an indicator of whether the individualperceives the governmental authorities and in-stitutions as responsive to citizen influence (seeCutler 2002).

Before ending this section, I want tosay something about results from parliamentsoutside the sphere of Western democracies.Rwanda, with a parliament of 48.8% women,deserves attention. Interviews with women rep-resentatives in the Rwandan parliament indi-cate that they consider themselves to be moreconcerned with grassroots politics than theirmale colleagues are, and also that there has beena strong advocacy of “international feminism”by many female deputies (Devlin & Elgie 2008).However, when Devline & Elgie examine pol-icy outputs, they conclude that the increasedwomen’s representation in Rwanda has hadlittle effect.

In summary, a large number of indicatorson substantive representation are used in em-pirical research. They are perhaps not ideal forcapturing the effects of the increased number ofwomen elected, but they are reasonable. Longi-tudinal studies comprise one way of advancingresearch in this field; however, what I believe iseven more urgent is the development of a setof standard definitions and indicators that en-able good cross-country comparative research.There is a trade-off here between the con-textual approach suggested above, which saysthat definitions of women’s interests and gen-der equality should be anchored in time andspace, and the comparison of a large number ofcases. However, even if theoretical definitionsshould end up a bit flat, I believe that compar-isons across countries are a necessary next step.Some indicators suit some contexts better than

64 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

other contexts; for example, registering legisla-tive voting behavior makes more sense in re-search on the U.S. Congress than in research onEuropean parliaments, where parties are morecoherent and pressure for loyalty is strong whenit comes to voting in the chamber (Lovenduski& Norris 2003). For European countries, it isnecessary to use indicators that capture impactsin earlier stages of the parliamentary process.

POLITICS OF PRESENCEOR FEMINIST AWARENESS?

The most interesting challenge to the theoryof the politics of presence is currently foundin the writings of, among others, Iris MarionYoung. This alternative approach highlights theimportance of feminist awareness. Instead of fo-cusing on the common experiences shared byfemale representatives and female voters (thepolitics of presence), the feminist awareness ap-proach concentrates on the formulation and im-plementation of programs explicitly aiming tochange society in women-friendly directions.The theory of feminist awareness does not as-cribe importance to female politicians per se,but to politicians with a feminist agenda (Young2000).

There is a tension in the literature on sub-stantive representation that has the potential tosharpen analyses on women in parliaments andmake the theory of the politics of presence moreprecise. It is obvious that not all women in par-liaments focus on women’s interests and genderequality, and it is also obvious that some menin parliaments are active in this field. However,I sense a risk that instead of cross-fertilization,there will be a split into two separate subfields:one focusing on women as voters and electedrepresentatives, and one focusing on women’smovements and “femocrats” in public adminis-tration.

During recent decades, governments inmost Western democracies have developeda set of agencies to meet the demands ofwomen’s movements. Research on state fem-inism labels these agencies women’s policyagencies (WPAs). Lovenduski (2005b, p. 4)

and colleagues suggest that “WPAs could in-crease women’s access to the state by further-ing women’s participation in political decision-making and by inserting feminist goals intopublic policy. Thus WPAs may enhance the po-litical representation for women.” Using sim-ilar reasoning, Weldon (2002, p. 1153) statesthat “women’s movements and women’s policyagencies may provide more effective avenuesof expression for women’s perspective than thepresence of women in the legislatures.” A lastexample from this strand of research can bedrawn from Sawer (2002, p. 17), who arguesthat increasing the number of women in parlia-ment is “insufficient” to ensure that women arebetter represented.

Research on institutions and actors outsidethe parliamentary process can help fill in thepicture of mechanisms that drive change in so-ciety. However, it has to be remembered thatthe number of female citizens who are part ofthe women’s movement is limited. It is also im-portant to remember that femocrats, in con-trast to female politicians, are seldom tried outin general elections. Although the parliamen-tary process should not be idealized, electionsare an important mechanism to check or cor-rect elites in society (Dahl 1989; Pitkin 1967,pp. 232–43). It would be a mistake for researchon substantive representation to sidestep theparliamentary process. From my point of view,a split into separate subfields would be a failurefor research on women and substantive repre-sentation.

FINAL COMMENTS

The result that emerges from the empiricalresearch is that female politicians contributeto strengthening the position of women’s in-terests. There is a need for more researchon women in parliaments, especially regardingsubstantive representation. The lack of cross-country comparative research has already beenhighlighted, but there is also a need for otherstudies, such as case studies exploring causalmechanisms in more detail. The questgoals areto see how, exactly, change takes place and to

www.annualreviews.org • Women in Parliaments 65

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

explore how the presence of women in par-liament might influence the behavior of men.Detailed studies would also be useful to moreclosely connect research on the causes of theincreased number of women elected and re-search on the effects of that change. There isan interplay between the two steps or levels ofthe parliamentary process—elections and pol-icy making—but the different sets of theoreti-cal and empirical research on women in parlia-ments are not always well integrated.

The list of desired studies is rather long. IfI were to single out one future developmentas more important than others, it would bethe quest for more cross-country comparativeresearch. Good-quality cross-country compar-ative empirical research is a sign of a mature re-search field. And a pending question in researchon women in parliaments is whether some par-liaments in the world are more open to changethan others.

The last theme I want to touch upon is aparadox that can be found between the lines inthis essay. Gender differences observed withinthe parliamentary process—such as gender dif-ferences in attitudes, priorities, and policypromotion—can be seen as a vehicle for change,even though segregation between women andmen elsewhere in society generally is seen asa mechanism to maintain the prevailing ordersof dominance and subordination (see Kimmel2004).

I believe it is almost impossible to go froma low proportion of women in parliament to ahigh proportion without going through a stagewherein visible divisions between female andmale politicians appear. However, the fact thatwomen at some phases concentrate on issuessuch as gender equality and social welfare doesnot necessarily mean that they should confinethemselves to these areas for all time. There is,naturally, a risk that the patterns described inthis overview, if maintained over long periods,conserve rather than change prevailing ordersof gender and power.

How should gender differences in the par-liamentary process be interpreted? A firm an-swer is impossible. Taking a bird’s-eye view,there are two perspectives on women in par-liaments, which can be denoted “static” and“dynamic.” From a static perspective, the pat-tern that emerges in this essay is interpretedas a division of labor between female and malepoliticians; when women enter the parliamen-tary arena, they take over certain areas frommen, but nothing becomes fundamentally dif-ferent as a result. The dynamic interpretationsees the emergence of genuine change. To me,the dynamic perspective is more credible. Thisjudgment does not rely on one or two espe-cially significant studies, but rather on the over-all result that emerges when numerous stud-ies covering a wide set of indicators are piledtogether.

DISCLOSURE STATEMENT

The author is not aware of any affiliations, memberships, funding, or financial holdings that mightbe perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

I thank Vicky Randall for her comments, especially concerning the section “Preconditions forChange.” I also thank Mette Anthonsen, Peter Esaiasson, and Mikael Gilljam for their helpfulcomments.

LITERATURE CITED

Atkeson LR, Carrillo N. 2007. More is better: the influence of collective female descriptive representation onexternal efficacy. Polit. Gender 3(01):79–101

66 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Banducci SA, Karp JA. 2000. Gender, leadership and choice in multiparty systems. Polit. Res. Q. 53(4):815–48Beckwith K. 2007. Numbers and newness. The descriptive and substantive representation of women. Can. J.

Polit. Sci. 40(1):27–49Bergqvist C, et al. 2000. Equal Democracies? Gender and Politics in the Nordic Countries. Oslo: Norwegian Univ.

PressBox-Steffensmeier JM, de Boef S, Lin TM. 2004. The dynamics of the partisan gender gap. Am. Polit. Sci.

Rev. 98(3):515–28Bratton KA, Ray LP. 2002. Descriptive representation, policy outcomes, and municipal day-care coverage in

Norway. Am. J. Polit. Sci. 46(2):428–37Celis K. 2006. Substantive representation of women: the representation of women’s interests and the impact

of descriptive representation in the Belgian parliament (1900–1979). J. Wom. Polit. Policy 28(2):85–114Childs S, Krook ML. 2006. Should feminists give up on critical mass? A contingent yes. Polit. Gender 2(4):522–

30Cutler F. 2002. The simplest shortcut of all: sociodemographic characteristics and electoral choice. J. Polit.

64(2):466–90Dahl R. 1989. Democracy and its Critics. New Haven, CT: Yale Univ. PressDahlerup D. 1988. From a small to a large minority: women in scandinavian politics. Scand. Polit. Stud.

4:275–98Dahlerup D, ed. 2006a. Women, Quotas and Politics. London: RoutledgeDahlerup D. 2006b. The story of the theory of critical mass. Polit. Gender 2(4):511–22Devlin C, Elgie R. 2008. The effect of increased women’s representation in parliament. The case of Rwanda.

Parl. Aff. 61(2):237–54Diaz MM. 2005. Representing Women? Female Legislators in West European Parliaments. Colchester, UK: ECPR

PressDodson DL. 2006. The Impact of Women in Congress. Oxford, UK: Oxford Univ. PressDollar D, Fisman S, Gatti R. 1999. Are women really the “fairer” sex? Corruption and women in government.

J. Econ. Behav. Org. 46:423–29Dietz MG. 2003. Current controversies in feminist theory. Annu. Rev. Polit. Sci. 6(1):399–431Dolan K. 2005. Do women candidates play to gender stereotypes? Do men candidates play to women? Can-

didate sex and issues priorities on campaign websites. Polit. Res. Q. 58:31–44Duverger M. 1955. The Political Role of Women. Paris: UN Econ. Soc. Counc.Esaiasson P. 2000. Focus of representation. See Esaiasson & Heidar 2000, pp. 51–82Esaiasson P, Heidar K, eds. 2000. From Westminster to Congress and Beyond. The Nordic Experience. Columbus:

Ohio State Univ. PressEsaiasson P, Holmberg S. 1996. Representation from Above. Members of Parliament and Representative Democracy

in Sweden. Aldershot, UK: DartmouthFreidenvall L. 2006. Vagen till Varannan damernas. Om kvinnorepresentation, kvotering och kandidaturval i svensk

politik 1970–2002 [The road towards every second seat for a woman. Women’s representation, quotas,and candidate selection in Swedish politics 1970–2002]. PhD thesis, Stockholm Stud. Polit.

Gender quotas I. 2005. Polit. Gender 1(4):621–52Gender quotas II. 2006. Polit. Gender 2(1):101–28Graubard SR, ed. 1986. Norden—The Passion for Equality. Oslo: Norwegian Univ. PressGrey S. 2002. Does size matter? Critical mass and New Zealand’s women MPs. In Women, Politics and Change,

ed. K Ross, pp. 19–29. Oxford, UK: Oxford Univ. PressGrey S. 2006. Numbers and beyond. The relevance of critical mass in gender research. Polit. Gender 2(4):492–

502Haavio-Mannila E, et al. 1983. Det uferdige demokratiet. Kvinner i nordisk politikk [Unfinished Democracy.

Women in Nordic Politics]. Oslo: Nordisk MinisterradHeath RM, Schwindt-Bayer LA, Taylor-Robinson MM. 2005. Women on the sidelines: women’s representa-

tion on committees in Latin American legislatures. Am. J. Polit. Sci. 49(2):420–36Heidar K, Pedersen K. 2006. Party feminism: gender gaps within Nordic political parties. Scand. Polit. Stud.

29(3):192–218

www.annualreviews.org • Women in Parliaments 67

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Hernes H. 1987. Welfare State and Woman Power: Essays in State Feminism. Oslo: Oslo Norwegian PressInglehart R, Norris P. 2003. Rising Tide. Gender Equality and Cultural Change Around the World. Cambridge,

UK: Cambridge Univ. PressIversen T, Rosenbluth F. 2008. Work and power. The connection between female labor force participation

and female political representation. Annu. Rev. Polit. Sci. 11:479–95Jeydel A, Taylor AJ. 2003. Are women legislators less effective? Evidence from the U.S. House in the 103rd–

105th Congress. Polit. Res. Q. 56(1):19–27Kenworthy L, Malami M. 1999. Gender inequality in political representation: a worldwide comparative anal-

ysis. Soc. Forces 78(1):235–68Kimmel M. 2004. The Gendered Society. Oxford, UK: Oxford Univ. PressKittilson MC. 2006. Challenging Parties, Changing Parliaments. Women and Elected Office in Contemporary Western

Europe. Columbus: Ohio State Univ. PressKoch JW. 2002. Gender stereotypes and citizens’ impressions of House candidates’ ideological orientations.

Am. J. Polit. Sci. 46(2):453–62Lovenduski J. 2005a. Feminizing Politics. Cambridge, UK: PolityLovenduski J, ed. 2005b. State Feminism and Political Representation. Cambridge, UK: Cambridge Univ. PressLovenduski J, Norris P, ed. 1993. Gender and Party Politics. London: SageLovenduski J, Norris P. 2003. Westminster women: the politics of presence. Polit. Stud. 51(1):84–102Matland RE. 1998. Women’s representation in national legislatures: developed and developing countries. Legis.

Stud. Q. 23(1):109–25Matland RE, Brown DD. 1992. District magnitude’s effect on female representation in the U.S. state legisla-

tures. Legis. Stud. Q. 17(4):469–92McAllister I, Studlar D. 1992. Gender and representation among legislative candidates in Australia. Comp.

Polit. Stud. 3:388–411Narud HM, Valen H. 2000. Does social background matter? See Esaiasson & Heidar 2000, pp. 83–106Norris P. 1993. Conclusions: comparing legislative recruitment. See Lovenduski & Norris 1993, pp. 309–30Norris P. 1996. Legislative recruitment. In Comparing Democracies. Elections and Voting in Global Perspective, ed.

L LeDuc, RG Niemi, P Norris, pp. 184–215. London: SageNorris P, Lovenduski J. 1995. Political Recruitment: Gender, Race and Class in the British Parliament. Cambridge,

UK: Cambridge Univ. PressOndercin HL, Bernstein JL. 2007. Context matters: the influence of state and campaign factors on the gender

gap in senate elections, 1988–2000. Polit. Gender 3(1):33–53Paxton P, Kunovich S. 2003. Women’s political representation: the importance of ideology. Soc. Forces 82(1):87–

113Pfau-Effinger B. 1998. Gender cultures and the gender arrangement—a theoretical framework for cross-

national comparisons on gender. Eur. J. Soc. Sci. 11(2):147–66Phillips A. 1995. The Politics of Presence. Oxford, UK: Oxford Univ. PressPhillips A. 2007. Multiculturalism without Culture. Princeton, NJ: Princeton Univ. PressPitkin HF. 1967. The Concept of Representation. Berkeley: Univ. Calif. PressPoggione S. 2004. Exploring gender differences in state legislators’ policy preferences. Polit. Res. Q. 57(2):305–

14Reignold B. 2000. Representing Women: Sex, Gender, and Legislative Behavior in Arizona and California. Chapel

Hill, NC: Univ. North Carolina PressRosenbluth F, Salmond R, Thies MF. 2006. Welfare works: explaining female legislative representation. Polit.

Gender 2(02):165–92Rule W. 1987. Electoral systems, contextual factors and women’s opportunity for election to parliament in

twenty-three democracies. West. Polit. Q. 40(03):477–98Sainsbury D. 1993. The politics of increased women’s representation: the Swedish case. See Lovenduski &

Norris 1993, pp. 263–90Salmond R. 2006. Proportional representation and female parliamentarians. Legis. Stud. Q. 31(2):175–204Sanbonmatsu K. 2003. Gender-related political knowledge and the descriptive representation of women. Polit.

Beh. 25(4):367–88

68 Wangnerud

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

ANRV377-PL12-04 ARI 13 April 2009 12:41

Sawer M. 2002. The representation of women in Australia: meaning and make-believe. In Women, Politics, andChange, ed. K Ross, pp. 5–18. Oxford, UK: Oxford Univ. Press

Schwindt-Bayer LA, Mishler W. 2005. An integrated model of women’s representation. J. Polit. 67(2):407–28Skjeie H. 1992. Den politiske betydningen av kjønn. En studie av norsk topp-politikk [The political importance of

gender. A study of Norwegian top-politics]. Oslo: Inst. Samfunnsforskning, rapport 92:11, (PhD thesis)Squires J, Wickham-Jones M. 2001. Women in parliament: a comparative analysis. Equal Opportunities Com-

mission, Res. Disc. Ser. (UK Gov. Publ.), 88–99Studlar DT, McAllister I. 2002. Does a critical mass exist? A comparative analysis of women’s legislative

representation since 1950. Eur. J. Polit. Res. 41:233–53Sung H. 2003. Fairer sex or fairer system? Gender and corruption revisited. Soc. Forces 82(2):703–23Swers ML. 1998. Are women more likely to vote for women’s issue bills than their male colleagues? Legis.

Stud. Q. 23(3):435–48Teigen M, Wangnerud L. 2009. Tracing gender equality cultures. Elite perceptions of gender equality in

Norway and Sweden. Polit. Gender 5(1): In pressThomas S. 1994. How Women Legislate. Oxford, UK: Oxford Univ. PressVega A, Firestone JM. 1995. The effects of gender on congressional behavior and the substantive representation

of women. Legis. Stud. Q. 20(2):213–22Weldon LS. 2002. Beyond bodies: institutional sources of representation for women in democratic policy-

making. J. Polit. 64(4):1153–74Wide J. 2006. Kvinnors politiska representation i ett jamforande perspektiv. Nationell och lokal niva [Women’s political

representation in a comparative perspective—national and local level]. PhD thesis, StatsvetenskapligaInst., Umea Universitet)

Wangnerud L. 1998. Politikens andra sida. Om kvinnorepresentation i Sveriges Riksdag [The second face ofdemocracy. Women’s representation in the Swedish parliament]. PhD thesis, Goteborg Stud. Polit.,Goteborg Univ.

Wangnerud L. 2000a. Testing the politics of presence. Women’s representation in the Swedish riksdag. Scand.Polit. Stud. 23(1):67–91

Wangnerud L. 2000b. Representing women. See Esaiasson & Heidar 2000, pp. 132–54Wangnerud L. 2006. A step-wise development: women in parliament in Sweden. In Women in Parliament:

Beyond Numbers, ed. J Ballington, A Karam, pp. 238–48. Stockholm: IDEA Handbook. Rev. versionYoung IM. 2000. Inclusion and Democracy. Oxford, UK: Oxford Univ. Press

www.annualreviews.org • Women in Parliaments 69

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

AR377-FM ARI 7 April 2009 11:25

Annual Review ofPolitical Science

Volume 12, 2009Contents

A Conversation with Robert A. DahlRobert A. Dahl and Margaret Levi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Neorepublicanism: A Normative and Institutional Research ProgramFrank Lovett and Philip Pettit � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �11

Domestic Terrorism: The Hidden Side of Political ViolenceIgnacio Sánchez-Cuenca and Luis de la Calle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �31

Women in Parliaments: Descriptive and Substantive RepresentationLena Wängnerud � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �51

Self-Government in Our TimesAdam Przeworski � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �71

Social Policy in Developing CountriesIsabela Mares and Matthew E. Carnes � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �93

Variation in Institutional StrengthSteven Levitsky and María Victoria Murillo � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 115

Quality of Government: What You GetSören Holmberg, Bo Rothstein, and Naghmeh Nasiritousi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 135

Democratization and Economic GlobalizationHelen V. Milner and Bumba Mukherjee � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 163

Has the Study of Global Politics Found Religion?Daniel Philpott � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 183

Redistricting: Reading Between the LinesRaymond La Raja � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 203

Does Efficiency Shape the Territorial Structure of Government?Liesbet Hooghe and Gary Marks � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 225

Bargaining Failures and Civil WarBarbara F. Walter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 243

Hobbesian Hierarchy: The Political Economy of PoliticalOrganizationDavid A. Lake � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 263

v

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.

AR377-FM ARI 7 April 2009 11:25

Negative CampaigningRichard R. Lau and Ivy Brown Rovner � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 285

The Institutional Origins of Inequality in Sub-Saharan AfricaNicolas van de Walle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 307

RiotsSteven I. Wilkinson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 329

Regimes and the Rule of Law: Judicial Independence in ComparativePerspectiveGretchen Helmke and Frances Rosenbluth � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 345

Field Experiments and the Political Economy of DevelopmentMacartan Humphreys and Jeremy M. Weinstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 367

Laboratory Experiments in Political EconomyThomas R. Palfrey � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 379

Field Experiments on Political Behavior and Collective ActionEline A. de Rooij, Donald P. Green, and Alan S. Gerber � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 389

Experiments on Racial Priming in Political CampaignsVincent L. Hutchings and Ashley E. Jardina � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 397

Elections Under AuthoritarianismJennifer Gandhi and Ellen Lust-Okar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 403

On Assessing the Political Effects of Racial PrejudiceLeonie Huddy and Stanley Feldman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 423

A “Second Coming”? The Return of German Political TheoryDana Villa � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 449

Group Membership, Group Identity, and Group Consciousness:Measures of Racial Identity in American Politics?Paula D. McClain, Jessica D. Johnson Carew, Eugene Walton, Jr.,and Candis S. Watts � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 471

Opiates for the Matches: Matching Methods for Causal InferenceJasjeet Sekhon � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 487

Indexes

Cumulative Index of Contributing Authors, Volumes 8–12 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 509

Cumulative Index of Chapter Titles, Volumes 8–12 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 511

Errata

An online log of corrections to Annual Review of Political Science articles may be foundat http://polisci.annualreviews.org/

vi Contents

Ann

u. R

ev. P

olit.

Sci

. 200

9.12

:51-

69. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by E

TH

- E

idge

noss

isch

e T

echn

isch

e H

ochs

chul

e Z

uric

h -

BIB

LIO

TH

EK

on

10/0

7/11

. For

per

sona

l use

onl

y.