Content uploaded by Elin Bjarnegård
Author content
All content in this area was uploaded by Elin Bjarnegård on Aug 30, 2021
Content may be subject to copyright.
Political Parties and Gender Quota Implementation
The Role of Bureaucratized Candidate Selection Procedures
Elin Bjarnegård and Pär Zetterberg
In recent years, the internal activities of political parties have increasingly become subject
to state regulations that prescribe their operations.
1
As a response to political parties’
inability to break male political dominance, almost eighty countries have adopted gender
quota laws (specified in the constitution or electoral code) that restrict the pre-electoral
choices that political parties make.
2
Such laws usually require parties to nominate a speci-
fied percentage of women in national elections.
3
A growing body of research on the
impact of compulsory quotas has shown that these quota laws may, but do not neces-
sarily, increase the number of female legislators.
4
One reason why even legally adopted
quotas are not always effective is that political parties charged with their implementation
do not comply with the quota requirements. Some parties simply ignore the quota;
others put a smaller number of women on the candidate lists than the law prescribes.
There are also parties that comply with the letter of the quota law but fail to implement
the spirit of the law (i.e. to get more women elected) by putting women at the bottom of
the candidate lists.
5
As a consequence, empirical research has shown that quota laws are
most effective whenever they include strong sanctions (e.g., nullification of the candi-
date list) for non-compliance with either the legally specified proportion of women on
the list or with placement mandates (that ensure that women are put in electable slots on
the lists).
6
A limitation of focusing only on quota design for explaining variations in the effi-
cacy of quota laws is that it implicitly portrays political parties as homogeneous,
responding equally to quota regulations. Yet, several studies emphasize that political
parties are heterogeneous, varying significantly in organizational structure, ideology,
candidate selection rules, party culture, etc.
7
Because parties are tasked with implement-
ing a legal reform that they did not necessarily wish for or design themselves, the dif-
ferences between political parties are potentially crucial for understanding the outcome
of the reform. Party factors shape political parties’responses to quota policies: some
parties are prone to implement new laws, such as quotas, whether there are enforcement
393
mechanisms in place or not, whereas others comply with regulations imposed on them
from the outside only when there are strong repercussions for not doing so. Hitherto,
however, the intraparty mechanisms that explain variation in quota implementation
remain largely unexplored.
8
Research has mostly shown that parties with a leftist ide-
ology tend to be better at complying with quota laws, and select more women candi-
dates, than other parties.
9
In this article, we focus on the organizational propensity of different types of
political parties to implement legislation. More specifically, the article pays attention
to one aspect of party organization, namely, the level of bureaucratization of the can-
didate selection process. A bureaucratized nomination process here refers to a process
that is carried out according to written rules that are “detailed, explicit, standardized,
implemented by party officials, and authorized in party documents.”
10
In other words,
bureaucratization refers to the rule-boundedness of a selection procedure, i.e., to whether
actual party decision-making reflects formal party rules. We draw on previous research
on gender and political recruitment to test the hypothesis that political parties with
bureaucratized candidate selection procedures are better at implementing legally man-
dated gender quotas than parties with largely informal selection procedures.
11
Until
now, the lack of adequate large-scale data at the party level has made it difficult to
properly operationalize the concept of bureaucratization of candidate selection and to
examine empirically its impact on quota implementation.
12
We fill this gap in the liter-
ature by testing the hypothesis using unique data covering almost 100 political parties in
eighteen Latin American countries. For each of the parties, the data include information
from the database Género y Partidos Políticos en América Latina [Gender and Political
Parties in Latin America] (GEPPAL) about issues such as the number of women on
the candidate lists, the number of women elected, how candidates are de facto selected,
party ideology, mechanisms to increase the representation of women and minorities,
etc.
13
We supplement these data, which were collected in 2009 by the Inter-American
Development Bank and International IDEA, with a textual analysis of the party statutes
for each of the parties in the GEPPAL dataset, in order to map out the official party
rules for candidate selection. We construct a measure of bureaucratized candidate selection
procedures by combining our collected data on formal party rules with the GEPPAL data
on de facto decision-making in processes of candidate selection.
We test the hypothesis using two models, each with a different specification of the
dependent variable. The first test assesses the impact of bureaucratized selection pro-
cedures on the implementation of the letter of the quota law by measuring the number
of women on each party’s candidate list in the last election prior to the GEPPAL data
collection. The second test also takes the spirit of the law into account and measures
implementation as the number of female representatives (by political party) in parlia-
ment following the same election.
To preview the results, we find partial support for the hypothesis: Whenever
quotas are in place, parties with bureaucratized selection procedures have on average
six percentage points, or 20 percent, more women candidates than political parties that
have informal selection procedures. However, while having bureaucratized recruitment
Comparative Politics April 2016
394
structures does increase the degree to which parties implement the letter of the quota
law, they do not necessarily generate implementation of the spirit of the law. Parties
with rule-bound selection procedures do not get more women elected after quota laws
are in place. Taken together, the analysis thus suggests that parties with bureaucratized
candidate selection procedures function as rule followers, but not as gender equalizers.
By performing the analysis, we bring together three bodies of literature that rarely
speak to each other: party literature, research on electoral quotas, and neo-institutional
theory. The analysis represents an initial attempt to properly operationalize the concept
of bureaucratization and, using large-scale statistical data on political parties across
countries, empirically examine its impact on quota implementation.
The Implementation and Effectiveness of Gender Quotas
There are various types of electoral gender quotas that apply to different stages of the
recruitment process (pre-election and election stage) and have different legal status
(voluntarily adopted by individual political parties or legally mandated in a country’s
constitution or electoral code). This study focuses on legally mandated gender quotas
that operate at the pre-election stage within the candidate recruitment structures of
political parties. In contrast to so-called reserved seats quotas, which prescribe a mini-
mum number of women among the elected representatives, quotas at the pre-election
stage require—by law—that all political parties have a minimum number (percentage)
of women among their candidates (so-called legislated candidate quotas). In a few
cases, the law instead targets the number of women contestants in party primaries
(so-called legislated aspirant quotas).
14
A fundamental requirement for quotas to be effective is that political parties choose
to implement the quota rule in the first place. However, research on the implementation
of legislated candidate quotas has shown that parties are often reluctant to do what
the law stipulates. The implementation of candidate quotas challenges power balances
within parties by effectively demanding that party gatekeepers put (female) newcomers
on the lists and remove some (male) candidates that were previously on the lists. Thus,
implementation of quotas is hindered by the very fact that it tends to generate resistance
within party organizations.
15
As a result of poor implementation, a number of states have revised their quota
laws and introduced strong sanctions, such as nullification of candidate lists, for
parties that fail to comply with the requirement to place the legally stipulated number
of women on their candidate lists.
16
Sanctions can also apply to more specific aspects
of the quota law. Putting the required number of women on the lists does not neces-
sarily make quotas effective, that is, it does not necessarily get more women elected.
Many political parties have chosen to comply with quotas in a minimal way by putting
the required number of women at the bottom of the candidate lists in non-electable
slots.
17
Therefore, a number of countries, such as Mexico, Argentina, Costa Rica, and
Bolivia, have also included sanctions that apply to rank-order requirements (or placement
Elin Bjarnegård and Pär Zetterberg
395
mandates) that stipulate where on the list the required number of women must be found.
Parties that fail to comply with the requirements are penalized.
Consequently, strong sanctions for parties that fail to put the required number of
women on the lists have become an important tool to guarantee that political parties
comply with the letter of the quota laws. Sanctions that also apply to the rank order
of candidates further restrict party gatekeepers’room for maneuver and tend to make
it more likely that the spirit of the quota law is also implemented and thus that it
increases the number of women elected.
18
Bureaucratized Candidate Selection Procedures and Quota Implementation
While the literature on quota implementation has largely had a structural perspective
on quota design, a small but growing body of literature has started to focus on politi-
cal parties and their different responses to quota policies.
19
Even in a system with
sanctions that apply to placement mandates, that is, where parties are required to have
both sexes represented among their top candidates, they can opt to consistently put
men first. In other words, the quota is implemented with considerable room for party
agency. Not surprisingly, the literature has noted that some types of parties are more
prone to comply with compulsory quotas than others, regardless of whether enforce-
ment mechanisms are in place or not. To reiterate, most scholarly attention has been
paid to party ideology: left-wing parties seem to comply with quota laws to a greater
extent than other parties.
20
Our analysis pays attention to another set of party characteristics, namely, how
the party is organized with respect to candidate selection. Various aspects of candidate
selection may potentially impact quota implementation, such as the level of decision-
making (centralized or decentralized), the inclusiveness of the selectorate, or the bureau-
cratization of the candidate selection procedure.
21
While all three of these aspects are
commonly referred to as crucial aspects of the candidate selection procedure, the first
two focus on the content of the process, whereas the third instead concerns its form.
22
In
this article, we pay attention to the form. We investigate whether it matters for quota
implementation if there is a connection or a disconnect between parties’formal rules
and the de facto selection processes. To reiterate, we use Norris’s oft-cited definition
of bureaucratization, which suggests that bureaucratized selection procedures are carried
out according to written rules that are “detailed, explicit, standardized, implemented by
party officials, and authorized in party documents.”
23
The definition corresponds well
to those central aspects of a broader Weberian ideal typical bureaucracy that stipulate that
in order to be bureaucratized, an action should be undertaken based on written regula-
tions that apply to a specifically defined area of activities.
24
It also fits nicely in a neo-
institutionalist framework: a bureaucratized candidate selection procedure is a formally
institutionalized procedure.
25
Bureaucratized recruitment structures are usually contrasted with informal procedures.
In line with a neo-institutionalist perspective, we recognize that there are significant
Comparative Politics April 2016
396
differences within the group of informal procedures. What unites them is that de facto
processes are not reflected in formal rules. This may be because formal party rules are
either brief and unspecific or simply not implemented. Instead, informal arrangements
and personal relationships trump written rules. In some weakly institutionalized cases
this means that “procedures may vary from one selection to another,”
26
thus rendering
them very unpredictable and arbitrary. In informally institutionalized procedures, how-
ever, political parties have a set of standard operating practices that are stable over time,
although not reflected in formal party documents. For the purposes of this study, a key
distinction should be made between bureaucratized (formally institutionalized) proce-
dures that rely on written rules and those procedures that are instead based on informal
practices (be they informally institutionalized or not).
27
Norris explicitly puts the issue of recruitment structures in relation to quota poli-
cies and presents the hypothesis that is tested in this article: Political parties that have
bureaucratized candidate selection procedures are likely to be better at implementing
legislated gender quotas than parties with largely informal selection procedures.
28
The main idea behind the hypothesis is that political parties with bureaucratized
candidate selection processes are used to being regulated by formal rules. Consequently,
when new formal rules are introduced, even from the outside, they effectively change
their practices. In contrast, in parties whose candidate selection is largely guided by
informally institutionalized procedures, new formal rules compete with established
informal institutions that are often deeply entrenched, difficult to change, and that some-
times trump formal rules.
29
Although bureaucratization in candidate selection has been part of the discourse
on candidate selection and quota implementation since the 1990s, the hypothesis has,
to our knowledge, never been tested empirically. One reason for the lack of empirical
tests is that cross-country comparative data at the party level (including information
about the sex of candidates) have been scarce. As a consequence, the concept of
bureaucratization of candidate selection processes has not been sufficiently elaborated
on nor properly operationalized. To enable a systematic large-N analysis, we build on
Norris to theoretically specify and operationalize the concept of bureaucratization.
30
Conceptualizing Bureaucratization in Candidate Selection
If we take Norris’sdefinition of bureaucratized selection procedures seriously, we would
expect that written rules should be found in party documents, the writing should be
detailed enough to be clearly understood, and the rules should be followed. We are thus
faced with a three-dimensional concept of bureaucratized candidate selection processes,
encompassing formalization, specification, and implementation. This three-dimensionality
has never been explicitly acknowledged, either theoretically or empirically.
The first criterion, formalization, distinguishes between those political parties that
have written rules for how to select their candidates and those that have no formal rules
whatsoever with respect to nominating candidates. In order for a political party to be
Elin Bjarnegård and Pär Zetterberg
397
characterized as having bureaucratized candidate selection procedures, a first condition
is that the party has written rules for candidate selection processes.
Second, specification implies that formal rules should not be overly general or vague
and that following these rules actually imposes important restrictions on the candidate
selection procedure. The defining features of the candidate selection process should
be specified. Hazan and Rahat point to four aspects of candidate selection that should
be considered in any theoretical framework analyzing such a procedure: eligibility cri-
teria for candidacy, the selectorate, decentralization (or level of decision making), and
appointment and voting system.
31
In other words, reading the written rules, one should
get an as clear as possible understanding of what the process is supposed to look like:
who can become a candidate, who selects the candidates, where this selection takes
place, and what kind of decision rule is applied. The process should be specified as
precisely as possible, give clear criteria, and leave little room for improvisation.
The third criterion concerns the implementation of the formal party rules that
regulate the candidate selection process. What is de jure authorized and specified in
party regulation should also matter for the de facto selection process. It is possible that
formal written rules exist and that they are quite specific in their wording, but are
disregarded by political actors. Research has pointed to the difficulties associated
with relying on formal regulations to analyze candidate selection processes. It is only
in certain kinds of parties that formal party regulations actually reflect what the can-
didate selection process looks like. In many other parties, actual practices depart from
written rules.
32
Our elaboration of Norris’sdefinition of bureaucratization of candidate selection
procedures thus suggests three criteria that all have to be fulfilled for a candidate selection
process to be considered bureaucratized: formalization, specification, and implementation.
We illustrate our conceptualization of bureaucratization of candidate selection in Figure 1.
With this conceptual scheme as a baseline, we can construct a valid measure of bureau-
cratized candidate selection procedures (see methods section for details).
Case Selection: Latin American Political Parties
Our empirical focus is concentrated on Latin American political parties. Ever since
Argentina in 1991 became the first country in the world to adopt legislated candidate
quotas, Latin America has been at the forefront of the quota debate. In no other part
of the world has such a large number of countries, in a relatively short period of time,
adopted this type of quota law. Candidate quotas became a paradigm for other countries
in the region, producing contagion effects from Argentina to other Latin American
countries, mainly in the second half of the 1990s. By 2009, eleven of the eighteen Latin
American countries included in this analysis had included gender quotas in their con-
stitution or electoral code.
33
There are a few factors that make Latin American political parties suitable for
empirical scrutiny. First, Latin American party systems are typically fairly volatile,
Comparative Politics April 2016
398
but there is a great deal of variation in terms of party-organizational characteristics in
general and in terms of the level of bureaucratization of the candidate selection pro-
cedures in particular.
34
Second, the Latin American countries share a number of char-
acteristics, including religion (Catholicism being the main religion), language (Spanish,
except Brazil), Spanish colonial history (again, except Brazil), and political system
(presidential system), that make them suitable for a so-called most similar systems
design.
35
This makes it possible from the outset to roughly control for a number of country
factors. Although it is always difficult to generalize from one region to another, we have
no reason to believe that parties should react differently to the introduction of electoral
gender quotas in Latin America than in other regions. We therefore cautiously suggest that
the results presented here should be relevant in other regions as well.
Data and Methodology
In order to test the hypothesis, we make use of a combination of sources. The main
source of data is Género y Partidos Políticos en América Latina [Gender and Political
Parties in Latin America] (GEPPAL), which is a large dataset including extensive infor-
mation on ninety-four political parties in eighteen countries, some of which have adopted
a national quota law.
36
To reiterate, GEPPAL is an initiative by the Inter-American
Figure 1 Conceptualizing Bureaucratization in Candidate Selection: an Illustration
FORMALIZED?
Are there written rules regarding candidate selection?
Yes No
SPECIFIED?
Are the written rules detailed, explicit and standardized?
Yes No INFORMAL
IMPLEMENTED?
Do the written rules guide the selection process?
Yes No
BUREAUCRATIZED
Elin Bjarnegård and Pär Zetterberg
399
Development Bank and International IDEA, and data were collected in mid-2009.
The dataset concerns the election prior to data collection and the preceding candidate
selection procedures. Information in the dataset includes issues such as party ideology,
measures to increase the number of women and minorities, how candidates are de facto
selected, and election results.
37
One of the main advantages with GEPPAL is that for
each political party it includes information about both the number of women on the can-
didate lists and the number of women elected to parliament. Such comparative data at
the party level are scarce. In addition to the GEPPAL data, we have collected our own
data on formal party rules to construct a measure of bureaucratized candidate selection
procedures. The formal party rules are combined with the GEPPAL data on de facto
decision-making in processes of candidate selection.
Our research strategy is to assess whether parties that have bureaucratized selection
procedures put more women on the ballot and got more women elected than other
parties, when a quota law is put in place. To measure the impact, our units of analysis
are made up by all the parties included in the GEPPAL dataset, and our key factor is the
interaction between bureaucratization and legislated quotas that target the pre-election
phase. The inclusion of parties in countries that have not adopted quota laws is crucial
in order to be able to distinguish the conditional impact of bureaucratization that our
hypothesis refers to from a direct impact. Their inclusion provides us with a point of
reference that we can use to ascertain that the effect of bureaucratization is, indeed,
contingent on the presence of a quota law.
38
We present two models, each with a dif-
ferent specification of the dependent variable. The first model deals with the implemen-
tation of the letter of the quota law: we investigate whether parties with bureaucratized
candidate selection procedures put more women on their candidate list than other
parties, once a quota law has been adopted. Here, our dependent variable is the number
of women on each party’s candidate list in the last election prior to the GEPPAL data
collection. In the second model, we test whether they are also more likely to implement
the spirit of the quotas than other parties, that is, whether, once a quota law is in place,
they also get more women elected. In this model, the dependent variable is the number
of female representatives, by party, in parliament in the same election.
39
GEPPAL mainly compiles data at the party level, so we use information from
the Quota Project to measure one of the two main independent variables that are
hypothesized to interact with each other: national quota laws at the pre-election phase
(bureaucratization being the other).
40
Where there was a national quota law in place
during the election included in GEPPAL, this is coded as “one,”and parties in coun-
tries where no such law existed get a “zero.”
41
It is important to note from the start that
there is indeed a variation within these categories that calls for explanation beyond the
mere existence of quota legislation.
42
Among the fifty-six parties that selected candi-
dates under some form of legislated pre-election quota, the number of women candi-
dates still ranged from 3 to 51 percent. Thus, even though we see that parties that
operate under a quota law in general tend not only to select more women candidates
(29 percent versus 20 percent in parties in countries with no quota law) and also send
more women to parliament (24 percent versus 12 percent), we can also conclude that
Comparative Politics April 2016
400
there is a large and important variation within the group of parties that operate under
a quota law.
Regarding the operationalization of bureaucratization of candidate selection pro-
cesses, we construct a variable that includes the three dimensions of the concept: for-
malization, specification, and implementation. Formalization is measured by assessing
whether the party statutes include any information about how candidates to parlia-
ment are selected. A party is given a “zero”if there is no such information and a “one”
if there is. Specification is measured by two of the four indicators mentioned above:
first, whether the party statutes establish who selects the candidates (yes or no), that
is, whether there is information about the size of the selectorate (ranging from individual
composition to open primary election) and, second, whether it is established where can-
didates are selected (yes or no), that is, at which level: national or subnational. Values
for specification thus range from zero to two, depending on the number of criteria each
party meets.
43
Finally, implementation takes into account not only the information from
party statutes but also from the GEPPAL dataset, which includes information about the
actual candidate selection procedures used in the last election prior to data collection.
44
If the information in the party statutes is the same as the information in GEPPAL about
actual selection procedures, with respect to the size of the selectorate and the level of
decision-making, the candidate selection procedure is considered implemented.
45
Thus,
for implementation, values also range from zero to two, depending on how many criteria
are met (see Appendix A for coding of variables).
46
To construct a variable measuring the level of bureaucratization of candidate selec-
tion, we specify in this first empirical test of the hypothesis a fairly inclusive set of con-
ditions based on the three dimensions. To be regarded as a bureaucratized candidate
selection procedure (and be given the variable value “one”), (i) a written party document
should authorize formal procedures for candidate selection, (ii) the writing should
specify one or two of the defining features of candidate selection (i.e., who and/or
where), and (iii) at least one of the two features should be implemented. All other
party selection procedures are regarded as informal (value “zero”) (see also Figure 1).
We have thus deliberately opted for a rather generous operationalization of a bureau-
cratized candidate selection process in that the implementation of one aspect of the
formal party rules concerning candidate selection (“who”or “where”) is enough to
code the entire process as bureaucratized. Thereby, we are able to test whether even
small steps towards a more rule-bound selection process are beneficial for the imple-
mentation of specific party-targeted national laws.
As for the descriptive statistics of the bureaucratization variable, they show that
there is considerable variation across the candidate selection processes of the different
political parties: 37 percent of the selection procedures are coded as “informal”and
63 percent are regarded as “bureaucratized.”This variation is often visible within indi-
vidual countries as well; however, in some countries there is quite a uniform pattern. For
instance, with a few exceptions, political parties in Argentina, Chile, and Uruguay seem
to have fairly informal selection rules.
47
Conversely, bureaucratized selection procedures
appear to be fairly common in political parties in Paraguay and Costa Rica.
48
Elin Bjarnegård and Pär Zetterberg
401
In order to properly assess the impact of bureaucratized candidate selection proce-
dures on quota implementation, we suggest an analytical model that takes some impor-
tant control variables into account. Any relationship between bureaucratization and
quota implementation might be the result of a factor that has an impact on both the
likelihood of bureaucratizing the candidate selection procedures and the number of
women on the candidate lists as well as among the elected representatives. Here, we
use two control variables at the party level (both included in the GEPPAL data) and
five control variables at the national level (for coding, see Appendix A). Starting with
the party factors, we control, first, for the possibility that party ideology has an impact
not only on the number of female candidates but also on the likelihood of bureau-
cratizing the selection process. Second, we control for quotas that are voluntarily
adopted by a specific political party (i.e., party quotas). A party that takes its party quota
seriously may not even have to implement a quota law: it is already a fait accompli.
49
As for the national-level factors, we control, first, for a country’s level of democ-
racy, using a Polity score.
50
It is likely that more democratic countries have a rela-
tively large number of parties that are internally democratic, and internal democracy
requires bureaucratization; in addition, democracy, after a certain threshold, is con-
ducive to women’s candidacy in general elections.
51
Second, we take into account a
country’s level of socio-economic development, using the Human Development Index
(HDI).
52
A socio-economically developed country is likely to bring more women into
the labor force and thus into public life. This might have an impact both on political
party organizations and on women’s propensity to run for political office.
53
Third, we
control for a politico-institutional variable: the electoral system. It has been widely
acknowledged that proportional representation, especially with closed lists (that do
not allow for preferential voting), is conducive to women’s political representation.
54
It is possible that election systems also impact the internal party organization and the
propensity of parties for having rule-bound selection procedures. GEPPAL includes
data on the electoral system that the parties operate under, and we single out closed list
proportional systems from all other electoral systems. Fourth, we control for aspirant
quotas. If the legislated quota targets contestants in party primaries instead of the com-
position of candidate lists, its impact on quota implementation (as measured here) is
only observed indirectly. Fifth and finally, we control for two sets of enforcement mech-
anisms: In the analysis of women candidates, we control for quotas that include strong
sanctions for non-compliance with the required proportion of women on the candidate
lists. These data also come from the Quota Project. To reiterate, sanctions have been
introduced in some cases when parties have put a smaller number of women on their can-
didate lists than the law prescribes. In Latin America, the adopted sanctions require parties
to comply fully with the letter of the law in order to be able to compete in the elections.
Such strong and tangible sanctions are generally effective in increasing the number of
women candidates. In the analysis of women elected, we add placement mandates to
the picture and control for quotas that include sanctions for non-compliance with rank-
order requirements of the candidates. The latter type of sanctions has been shown to be
important for increasing the number of women elected to parliament.
Comparative Politics April 2016
402
A limitation in the empirical analysis is its use of cross-sectional data from a single
election. Ideally, we would need time-series data to address issues of causality. Such
data, however, do not exist in cross-national research on political parties’candidate
selection procedures and women’s candidacy in public elections. Nevertheless, we
use “lagged”control variables at the national level (i.e., explanatory factors measured
in 2005, at a point in time prior to the last election before the GEPPAL data collection).
By lagging the control variables, we can design our model so that it accounts for the time
order inherent in our causal argument.
The analytical model is performed using ordinary least squares (OLS) regression
analysis. There are a few limitations with OLS regression in analyses that have data
at different levels of analysis (here, party level and national level). First, the assumption
that observations (in this case, political parties) are independent from each other is likely
to be violated. Thus, we need to be aware of potential auto-correlation problems. Second,
the error terms are not likely to be randomly distributed across an explanatory factor;
that is, problems with heteroscedasticity are probable. We take the former problem into
account by treating only political parties from different countries as independent from
each other by clustering the standard errors. The second limitation is overcome by includ-
ing robust standard errors in the analysis.
Results
Model 1 in Table 1 presents our first test of the hypothesis that political parties with
bureaucratized candidate selection are likely to implement legislated quotas to a larger
degree than political parties that select their candidates in a more informal manner.
Model 1 generates support for the hypothesis. The interaction term used for assess-
ing the interaction effect between quotas and bureaucratization on the number of
women candidates is positive, fairly large, and significant at a 99 percent level of
significance. This indicates that bureaucratization of selection procedures does matter
for quota implementation.
55
The results presented in Model 1 in Table 1 have been subject to several robustness
tests, and they are insensitive to model specification: thus, additional analyses (not
reported here) confirm the results.
56
We have, for instance, specified the main inde-
pendent variable, bureaucratization, by using an index (ranging from 0 to 2) to specify
it as a graded scale rather than as a dichotomy.
A key follow-up question concerns the size of the conditioned impact of bureau-
cratized candidate selection procedures. Figure 2 is based on the results of the analysis
presented in Model 1. When we concentrate all other values in Model 1 to the mean
(or mode, whenever there is a categorical variable),
57
political parties with bureaucra-
tized selection procedures are expected to have candidate lists comprising 37.8 percent
women whenever a quota law is in place. For parties with informal selection proce-
dures, the expected proportion of women after a quota law is adopted is 31.5 percent.
Thus, there is a difference of six percentage points, or 20 percent. With the relatively
Elin Bjarnegård and Pär Zetterberg
403
small number of observations (<100), it is not possible to establish that the difference
between parties with bureaucratized selection procedures and those with informal
recruitment structures is statistically significant (the confidence intervals are huge,
and thus overlap). However, the difference in the number of women on the lists
between the two groups is substantial and is therefore unlikely to be the result of
mere chance. For this reason, we suggest that the findings point to a clear trend among
Latin American parties.
The adoption of a quota law appears to generate quite different reactions within
different political parties. For parties with a bureaucratized selection procedure, quota
adoption results in almost double the amount of women candidates, from 20 to 37.8 per-
cent (an 88 percent increase). On the other hand, quota adoption seems to generate much
Ta b l e 1 The Interaction Effect of Legislated Quotas and Bureaucratized Candidate
Selection Procedures On Women Candidates And Elected Women: Multivariate
Analysis
Model 1
Women Candidates
Model 2
Women Elected
Bureaucratization x Legislated quota 12.12
(4.12)
*** 3.40
(6.56)
Bureaucratization −5.78
(2.94)
*−2.46
(3.30)
Party ideology −0.83
(0.60)
−2.56
(1.55)
Party quota 3.94
(3.13)
−0.76
(5.70)
Democracy −3.65
(3.02)
−1.95
(1.87)
Socio-economic development 24.75
(27.27)
1.04
(28.43)
PR system with closed lists 14.71
(4.94)
*** 7.67
(5.26)
Legislated quota 5.58
(7.63)
9.18
(8.33)
Sanctions (list) 2.68
(7.49)
Sanctions (rank order) 11.83
(7.42)
Aspirant quota −14.95
(8.59)
2.90
(9.94)
Constant 27.72 33.66
R² (N) 0.43 (73) 0.27 (79)
Notes: Unstandardized OLS regression coefficients, (clustered) robust standard error in parenthesis. Checked
by VIF statistics to be free of multicollinearity problems. Missing 5list-wise deletion. Coding of variables
is listed in Appendix A.
*** 5sign. at < 0.01; ** 5sign. at < 0.05¸* 5sign. at < 0.10.
Comparative Politics April 2016
404
less of a change in how parties with a largely informal selection process operate. These
parties have a mere 21.5 percent increase in women candidates, from 25.9 to 31.5 percent.
Thus, having bureaucratized candidate selection procedures appears to be a beneficial
factor for women who run in internal party races for candidacy, however, only in con-
junction with the adoption of a law that aims at leveling out gender inequalities in
political candidacy. In contrast to what some authors have claimed, bureaucratization
in itself is not beneficial for women in candidate recruitment processes.
58
As shown
in Figure 2, bureaucratized parties have slightly lower levels of women candidates than
parties with informal selection procedures whenever there is no quota law in place.
59
The
parties investigated thus differ in their propensity to implement a law rather than in their
predisposition to include women.
To interpret the findings, we suggest that parties that already have bureaucratized
candidate selection procedures are used to following an established set of formal rules
when nominating candidates for election. However, there is an important difference
between internal party regulations and a quota law: the latter is imposed on the party,
whereas regulations are usually adopted by the party. More resistance could be expected
with regards to obeying national laws as compared to following one’sownformal
regulations, particularly if the law imposed is perceived as reducing the prospects of
electoral success. When a new piece of legislation, such as the quota law, is passed,
it is targeted towards altering the manner in which political parties go about selecting
their candidates. Although there is probably a certain level of resistance to imposed
rules in all parties, bureaucratized parties have a stronger organizational predisposition
to alter de facto processes based on changes in formal regulations. In parties with less
Figure 2 The Expected Number of Women Candidates (in Percent) as a Function of
Bureaucratization of Candidate Selection Procedures and Legislated Quotas
Elin Bjarnegård and Pär Zetterberg
405
bureaucratized candidate selection procedures, informally institutionalized behavior
may well guide political recruitment in predictable ways. However, because informally
institutionalized practices are based on established understandings of “how things are
done,”their logic is very difficult to break, and they tend to be very stable. As a con-
sequence, they often remain unaltered even in light of formal rule change.
60
Our find-
ings are thus consistent with case-study research on quota implementation that has
highlighted the difficulties of reforming recruitment due to “sticky”informal institu-
tional legacies. The fact that political actors often use informal practices to circumvent
formal rules and reforms, such as gender quotas, may account for the smaller number
of women candidates in parties with informal candidate selection processes.
61
Thus, after showing in our first test that political parties with bureaucratized can-
didate selection are more likely than other parties to comply with the letter of the law,
we want to see, in a second test, whether parties with these procedures are also more
likely to implement the spirit of the quotas. This would mean that bureaucratized parties
not only put the required number of women somewhere on the list, but that they put
them in electable slots, to ensure that the quota law will impact representation.
The continued analysis presented in Model 2 in Table 1, however, demonstrates
that the implementation of a quota law does not automatically imply that it has the effect
that policy makers intended. An interaction model with bureaucratization and legislated
quotas, in which the percentage of women elected (measured by party) is the dependent
variable, shows that there is no conditional impact of bureaucratization on the number
of women elected to parliament once a quota law is in place. As Model 2 shows, the
interaction term is no longer statistically significant, and the regression coefficient is
relatively small. The model is robust for various alternative model specifications.
62
In sum, we find that the hypothesis is given partial support: Political parties with
bureaucratized candidate selection put more women on their candidate lists; however,
they do not get more women elected. Our interpretation of this finding is that parties
with bureaucratized selection procedures put the right number of women on their lists
simply because they are required to do so. These parties have developed a set of routines
for implementing formal rules that also come into play when rules are imposed by the
government. This, however, does not mean that party officials take the spirit of the
quota rules into consideration; as a consequence, they do not necessarily put women
in the electable slots and get women into office. In other words, the fact that these
parties follow formal rules, obey laws, and implement requirements does not auto-
matically make them more gender equal.
Because all political parties strive to minimize internal power struggles, our analy-
sis suggests that there is a clear limit to the expectations we can have even on parties
with bureaucratized selection procedures. Whereas an implementation of the letter of
the quota laws may not instigate serious internal struggles, an implementation where
incumbent male candidates in electable slots are replaced with female candidates almost
certainly will.
63
In times of candidate nomination, political parties have more issues on
their agenda than the effective implementation of a quota law, including the running of
a successful campaign, rewarding loyal party candidates, maintaining peace among
Comparative Politics April 2016
406
party factions, keeping important alliances intact, pleasing voters, and, ultimately, win-
ning the election. For parties with bureaucratized candidate selection procedures, these
issues will perhaps not prevent them from following the new rules imposed by the
legislated quotas, but they may very well prevent them from doing more than the
minimum required of them.
Conclusion
This article has analyzed the relationship between political parties and gender quota
implementation. More specifically, we have tested the hypothesis that political parties
with bureaucratized candidate selection procedures are likely to be better at imple-
menting legislated quotas than parties with informal candidate selection procedures.
A prerequisite for this analysis is the improved conceptual clarity regarding bureaucra-
tization that this article contributes. Bureaucratization has sometimes been confounded
with related concepts such as institutionalization
64
and formalization,
65
but we argue
that formalization, specification, and implementation all need to be considered when
defining and measuring bureaucratization.
Our analysis provides partial support to the hypothesis. First, it shows that when-
ever there is a quota law in place, parties with bureaucratized selection procedures have
on average six percentage points (or 20 percent) more women candidates than political
parties with informal or informally institutionalized selection procedures. To understand
this finding, we point at the organizational predisposition to implement different types
of selection reforms that parties with bureaucratized candidate selection procedures are
likely to have. Candidate selection in these parties is dictated by formal rules, and when
the legal framework changes, affecting the internal rules of the party, actual practice
changes as a consequence, even if there is resistance to the new law as such. In informal
selection processes, on the other hand, formal rules do not have as large an impact on
the actual behavior of party actors. Instead of being regulated by written rules, party
practices for selecting candidates are based on established routines and deep-rooted
expectations. Party actors resisting reform thus have more leeway in parties with infor-
mal selection processes. More broadly, party culture seems to be strong in the sense that
parties are prone to continue behaving in much the same way as they always have. In
bureaucratized parties this means that they continue to follow formal rules, even when
the formal rules change. More informal parties, on the other hand, continue to depend
on unwritten practices and are thus prone to disregard new legal frameworks.
Second, we qualify the picture by demonstrating that implementation of the letter
of the quota law does not equal implementation of the spirit of the quotas. Parties with
rule-bound selection procedures do not get more women elected than other parties do.
Taken together, the findings indicate that political parties with bureaucratized selec-
tion procedures indeed are rule-followers, but that does not automatically make them
gender equalizers. This finding clearly suggests that bureaucratized selection pro-
cedures are no panacea for women in politics and that the expectations that part of
Elin Bjarnegård and Pär Zetterberg
407
the literature on gender and candidate selection has had on bureaucratization may have
been too high.
66
Giving due attention to political parties as organizations helps us understand why
the intentions of policy makers are not always fulfilled. The analysis suggests that in
order to fully understand when gender quotas, as well as other party-targeted reforms,
are complied with—and when they are not—researchers should pay specific attention to
the inner life of political parties, including both formal and informal aspects and the
interplay between them.
67
For gender quotas in particular, this insight is important from
a policy perspective as well: On the one hand, strengthening the organization of political
parties in terms of bureaucratizing their candidate selection processes seems to be an
effective way to ensure that parties meet the formal requirements of the quota policies.
On the other hand, this analysis has also brought to our attention that having such pro-
cedures does not ensure that they meet the spirit of the law and get more women into
office. Studying the details of internal party decision-making, including their gendered
aspects, is thus crucial to unveiling the possibilities as well as the limitations of this
global electoral reform.
NOTES
This article is based on work supported by the Riksbankens Jubileumsfond [grant no. P10-0801:1]. The
authors thank Amanda Clayton, Lenita Freidenvall, Mona Lena Krook, Martin Lundin, Petra Meier, Pippa
Norris, Jennifer Piscopo, Leslie Schwindt-Bayer, Eleonora Stolt, Michelle Taylor-Robinson, Tània Verge,
Thomas Widenstjärna, and three anonymous reviewers for helpful comments and suggestions. A special
thanks to Camilla Reuterswärd for excellent research assistance. Any remaining errors are the responsibility of
the authors. The authors are equal contributors to this article; names are listed in alphabetical order.
1. Ingrid van Biezen and Ekaterina R. Rashkova, “Deterring New Party Entry? The Impact of State
Regulation on the Permeability of Party Systems,”Party Politics, 20 (September 2014), 890–903.
2. Quota Database, http://www.quotaproject.org/, accessed November 21, 2015. There are also quota
laws that apply directly to the election results, so-called reserved seats. Reserved seats quotas, which prescribe
that a certain number of seats in parliament are reserved for women, are also increasingly common, but
not in the empirical context of interest in this study: Latin America. For an overview of different quota
types, see e.g., Mona Lena Krook, Quotas for Women in Politics: Gender and Candidate Selection Reform
Worldwide (New York: Oxford University Press, 2009).
3. Drude Dahlerup, ed., Women, Quotas and Politics (New York: Routledge, 2006). In a few cases, quota
laws instead require political parties to have a certain number of women among the contestants in internal
races for candidacy (so-called aspirant quotas).
4. Krook.
5. For illustrations of how political parties have operated in order to thwart effective implementation of
the quota laws, see e.g., Magda Hinojosa, Selecting Women, Electing Women: Political Representation and
Candidate Selection in Latin America (Philadelphia: Temple University Press, 2012).
6. Mala Htun and Mark P. Jones, “Engendering the Right to Participate in Decisionmaking: Electoral
Quotas and Women’s Leadership in Latin America,”in Nikki Craske and Maxine Molyneux, eds., Gender
and the Politics of Rights and Democracy in Latin America (London: Palgrave, 2002), 32–56; Mark P. Jones,
“Gender Quotas, Electoral Laws, and the Election of Women: Evidence from the Latin American Vanguard,”
Comparative Political Studies, 42 (November 2009), 56–81.
7. See e.g., Maria Escobar-Lemmon and Michelle M. Taylor-Robinson, “How Do Candidate Recruitment
and Selection Processes Affect Representation of Women?”in Peter Siavelis and Scott Morgenstern, eds.,
Pathways to Power (University Park: Pennsylvania State University Press, 2008); Rainbow Murray, “How
Comparative Politics April 2016
408
Parties Evaluate Compulsory Quotas: A Study of the Implementation of the ‘Parity’Law in France,”Parliamentary
Affairs, 60 (October 2007), 568–84; Tània Verge, “Gendering Representation in Spain: Opportunities and
Limits of Gender Quotas,”Journal of Women, Politics & Policy, 31 (2010), 166–90; Hinojosa.
8. See also Meryl Kenny, Gender and Political Recruitment: Theorizing Institutional Change (Houndmills:
Palgrave Macmillan, 2013).
9. See e.g., Murray.
10. Pippa Norris, “Legislative Recruitment”in Lawrence Le Duc, Richard G. Niemi, and Pippa Norris,
eds., Comparing Democracies: Elections and Voting in Global Perspective (London: Sage Publications,
1996), 184–215, 202. The concept of bureaucratization has sometimes been confounded with formalization
(see e.g., Elin Bjarnegård and Pär Zetterberg, “Removing Quotas, Maintaining Representation: Overcoming
Gender Inequalities in Political Party Recruitment,”Representation, 47 (2011), 187–99) and institutionaliza-
tion (see e.g., Kathleen Bruhn, “Whores and Lesbians: Political Activism, Party Strategies, and Gender
Quotas in Mexico,”Electoral Studies, 22 (March 2003), 101–09).
11. See e.g., Norris; Richard E. Matland, “Enhancing Women’s Political Participation: Legislative Recruit-
ment and Electoral Systems,”in Julie Ballington and Azza Karam, eds., Women in Parliament (Stockholm:
International IDEA, 2005).
12. Cf. Escobar-Lemmon and Taylor-Robinson.
13. GEPPAL, www.iadb.org/research/geppal, accessed July 14, 2011.
14. To the best of our knowledge, research on the very few instances of legislated aspirant quotas is virtually
non-existent. For an exception, see Tricia Gray, “Quota Mechanics in Panamá, 1999–2014: ‘Se obedece pero no
se Cumple’,”Bulletin of Latin American Research, 34 (July 2015), 289–304.
15. Hinojosa.
16. See e.g., Htun and Jones.
17. See e.g., Htun and Jones; Leslie Schwindt-Bayer, “Making Quotas Work: The Effect of Gender Quota
Laws on the Election of Women,”Legislative Studies Quarterly, 34 (February 2009), 5–28.
18. Ibid.
19. See e.g., Escobar-Lemmon and Taylor-Robinson; Murray; Hinojosa; Kenny; Bjarnegård and Zetterberg.
20. See e.g., Hinojosa, Murray.
21. See e.g., Pippa Norris and Joni Lovenduski, Political Recruitment. Gender, Race and Class in British
Parliament (Cambridge: Cambridge University Press, 1995); Hinojosa.
22. See e.g., Norris; Bonnie N. Field and Peter Siavelis, “Candidate Selection Procedures in Transitional
Politics: A Research Note,”Party Politics, 14 (September 2008), 620–39.
23. Norris, 202.
24. See e.g., Richard Swedberg and Ola Agevall. The Max Weber Dictionary: Key Words and Central
Concepts (Stanford: Stanford University Press, 2005), 19.
25. See also Flavia Freidenberg and Steven Levitsky, “Informal Institutions and Party Organizations in
Latin America,”in Gretchen Helmke and Steven Levitsky, eds., Informal Institutions and Democracy:
Lessons from Latin America (Baltimore: The Johns Hopkins University Press, 2006).
26. Norris, 203.
27. See also Freidenberg and Levitsky, 180–81 for a similar distinction. Also note that in studies with other
purposes than ours, it may be more relevant to make a key distinction between institutionalized (formally and
informally) and largely informal procedures. See e.g., Siavelis and Morgenstern.
28. Norris.
29. Elin Bjarnegård, Gender, Informal Institutions and Political Recruitment: Explaining Male Dominance
in Parliamentary Representation (Houndmills: Palgrave Macmillan, 2013); Kenny; Gretchen Helmke and
Steven Levitsky, “Informal Institutions and Comparative Politics: A Research Agenda,”Perspectives on Politics,
2 (December 2004), 725–40.
30. Norris.
31. Reuven Y. Hazan and Gideon Rahat, Democracy within Parties: Candidate Selection Methods and
Their Political Consequences (New York: Oxford University Press, 2010).
32. E.g., Michael Gallagher and Michael Marsh. Candidate Selection in Comparative Perspective: The
Secret Garden of Politics (London: SAGE, 1988); Norris; Hazan and Rahat; Bjarnegård.
33. Additional countries have adopted quotas since, but as these laws were implemented after the data
collection, they are disregarded here. See Appendix B for more details about the quota laws analyzed here,
available in the online version, at www.ingentaconnect.com/cuny/cp.
34. E.g., Manuel Alcántara Saez, Instituciones O Máquinas Ideológicas? Origen, Programa Y Organización
De Los Partidos Políticos Latinoamericanos (Barcelona: Instituto de Ciències Politiques i Socials, 2004).
Elin Bjarnegård and Pär Zetterberg
409
Note, however, that although Alcántara Saez categorizes a large number of Latin American parties on various
dimensions, he does not focus on the level of bureaucratization of the candidate selection process.
35. Adam Przeworski and Henry Teune, The Logic of Comparative Social Inquiry (New York: John Wiley &
Sons, 1970).
36. GEPPAL includes information about the five largest parties in each country’s lower chamber and about
all parties that obtain at least 5 percent representation. See GEPPAL and Appendix B for more information about
the countries included in the analysis. It should be noted that the GEPPAL dataset is incomplete with respect to
information about women candidates and representatives. For example, information about women candidates is
missing for all political parties of the Dominican Republic. In addition, there is missing data on some of the
control variables. This leaves us with information about seventy-three parties in seventeen Latin American coun-
tries for candidates and seventy-nine parties in eighteen Latin American countries for representatives.
37. To make sure that the information is accurate, the data collection process in each country was headed by
a country expert on gender and political representation and/or on political parties. The expert-led data
collection process reduces a common problem in comparative analyses on political parties: the researcher’s
lack of detailed knowledge about each country’s political system as well as about intra-party processes. See
Gallagher and Marsh.
38. See also Escobar-Lemmon and Taylor-Robinson, who use a similar approach (although their analysis
is mostly descriptive due to the limited number of cases).
39. The measure of female representatives is not perfect as an indicator of political parties’implemen-
tation of quotas, because voter behavior may have an impact on who gets elected. As an additional robust-
ness check, we have run an additional model with the number of women as top candidates on the list as the
dependent variable. The results from these analyses are not substantially different from the ones we present
in the results section.
40. Quota Project, “Global Database of Quotas for Women,”www.quotaproject.org (International IDEA and
Stockholm University. Accessed May 23, 2013).
41. See Appendices A and B for a more detailed description.
42. See Appendix C for descriptive statistics at the party and country level.
43. Note that when constructing the bureaucratization variable we are only interested in assessing whether
or not party statutes stipulate who selects candidates and where candidate selection takes place. We are only
interested in whether there are rules regulating this, and we do not focus on the content of those rules, i.e., who
these specific actors are or at which specific level the final decisions are made. This is in contrast to, for
instance, Hinojosa, who pays explicit attention to these two factors.
44. Information about the actual selection procedure is a result of national party experts’interviews with
party officials, exclusively conducted for the GEPPAL project. It should be mentioned that party officials might
have an interest in wrongfully saying that the selection procedure was conducted in line with what is
mentioned in the statutes, in order to put the party in a favorable light. However, we see this as a minor
problem, as exaggerating the level of bureaucratization is likely to lead to an underestimation of its impact
on quota implementation.
45. Note that for level of decision making, the matching of party statutes and party official is fairly difficult
whenever primaries are involved. Party officials do not mention the level at which primaries take place.
Therefore, we make an assumption that there is implementation in those cases in which both the statutes
and party official mention that primaries are used.
46. Due to space constraints, the Appendix is not in the print version of this article. It can be viewed in the
online version, at www.ingentaconnect.com/cuny/cp.
47. See also Juan Andrés Moraes, “Why Factions? Candidate Selection and Legislative Politics in Uruguay,”
in Siavelis and Morgenstern, eds., 164–85.
48. See also Mark P. Jones, “Quota Legislation and the Election of Women: Learning from the Costa Rican
Experience,”The Journal of Politics, 66 (November 2004), 1203–23.
49. See e.g., Caul Kittilson; however, see Hinojosa.
50. Polity IV Project, http://www.systemicpeace.org/polity/polity4.htm, accessed February 20, 2011.
51. Kathleen M. Fallon, Liam Swiss, and Jocelyn Viterna, “Resolving the Democracy Paradox: Democratiza-
tion and Women’s Legislative Representation in Developing Nations, 1975 to 2009,”American Sociological
Review, 77 (June 2012), 380–408; Elin Bjarnegård and Erik Melander, “Disentangling Gender, Peace and
Democratization: The Negative Effects of Militarized Masculinity,”Journal of Gender Studies,20(2011),139–54.
52. Human Development Index (HDI), UNDP, http://hdr.undp.org/, accessed February 20, 2011.
53. We considered including a measure of gender equality instead of or in addition to socio-economic
development. However, indices such as the Gender Inequality Index (GII) include our dependent variable:
Comparative Politics April 2016
410
women in politics. Separate measures of gender equality, such as fertility rates, male to female ratio of
education, and participation in the labor market do not impact on the main results.
54. Wilma Rule, “Electoral Systems, Contextual Factors and Women’s Opportunity for Election to
Parliament in 23 Democracies,”Western Political Quarterly, 40 (September 1987), 477–98. See also Jones.
55. The direction of the coefficients of the control variables is in line with our expectations from pre-
vious research.
56. As robustness checks, we first ran models in which we dropped control variables. We also ran models
with alternative measurements for some of the control variables. The analysis is insensitive to alternative
measurements of democracy and societal development. The significant conditional impact of bureaucratization
remains when introducing factors such as district magnitude, federalism, and age of the party.
57. Modes for the categorical variables are “no party quota,”“PR with closed lists,”“no sanctions,”and “no
aspirant quota.”
58. See e.g., Moshe M. Czudnowski, “Political Recruitment,”in Fred I. Greenstein and Nelson W. Polsby,
eds., Handbook of Political Science, (Reading: Addison Wesley, 1976) 155–242; Miki Caul Kittilson,
Challenging Parties, Changing Parliaments: Women and Elected Office in Contemporary Western Europe
(Columbus: The Ohio State University Press, 2007); Maria Guadagnini, “A‘Partiocrazia’without Women:
The Case of the Italian Party System,”in Joni Lovenduski and Pippa Norris, eds., Gender and Party
Politics (London: Sage, 1993); Matland.
59. The lack of an independent significant effect of bureaucratization has been confirmed in a separate
model where the interaction term has been excluded (not reported here).
60. Bjarnegård; Hans-Joachim Lauth, “Informal Institutions and Democracy,”Democratization, 7 (2000),
21–50.
61. See also Kenny. It should be added, however, that parties with informal candidate selection processes
are not completely unaffected by legal changes (see Figure 2). For instance, there may be debates within some
parties about the need to renew the pool of candidates in order to attract new voters. In such cases, quota
reforms partly overlap with, and give fuel to, already ongoing, informal discussions.
62. For instance, the results are similar when we run Model 2 including the exact same cases as in Model 1
(n573). This indicates that the different results in the two models are not driven by a few political parties. As
for the control variables, sanctions that apply to placement mandates are significant in alternative model
specifications (with fewer control variables), which is in line with previous research. However, it does not
have an impact on the main results.
63. See e.g., Marion Reiser, “The Universe of Group Representation in Germany: Analysing Formal and
Informal Party Rules and Quotas in the Process of Candidate Selection,”International Political Science
Review, 35 (January 2014), 55–66.
64. See e.g., Bruhn.
65. See e.g. Bjarnegård and Zetterberg.
66. Cf. Czudnowski; Guadagnini; Matland; Caul Kittilson.
67. Cf. Louise Chappell and Georgina Waylen, “Gender and the Hidden Life of Institutions,”Public
Administration, 91 (September 2013), 599–615.
Elin Bjarnegård and Pär Zetterberg
411
APPENDIX
Appendix A Coding of Variables
Women candidates: The number, in percent, of women on the party’s candidate lists
to the lower/single house in the last election prior to 2009. Source: GEPPAL.
Women elected representatives: The number, in percent, of each party’sfemaleelected
representatives to the lower/single house in the last election prior to 2009. Source:
GEPPAL.
Formalization: Whether a party has formal rules (in the party statutes) for the selection
of candidates to the national parliament. 1 5yes; 0 5no. Source: Authors’analysis of
party statutes.
Specification 1—selectorate size: Whether the selectorate that makes the final formal
decision in the candidate selection process is specified in the party statutes. 1 5yes;
05no. Source: Authors’analysis of party statutes.
Specification 2—decentralization: Whether the level at which the final decision of the
candidate selection process is made is specified in the party statutes. 1 5yes; 0 5no.
Source: Authors’analysis of party statutes.
Specification index: Additive index (0–2) based on Specification 1 and Specification 2.
Implementation: Whether the selection rules stated in the party statutes coincide with
the rules that the interviewed party official says were used in the last election prior to
2009. Applies to both selectorate size (variable values ranging from individual composi-
tion to open primary elections; cf. Hazan and Rahat, 2010) and level of decision-making
(variable values are national or sub-national level). Additive index (0–2) based on both
aspects of specification, for which 1 5do coincide and 0 5do not coincide. Source:
Authors’analysis of party statutes and GEPPAL.
Bureaucratization: Variable based on formalization, specification (index), and imple-
mentation. 0 5informal and 1 5bureaucratized (see main text for a closer description).
Party ideology: Score from 1 (left) to 10 (right), as perceived by party official. Source:
GEPPAL.
Party quota: Presence of voluntarily adopted party candidate quota. 1 5yes, 0 5no.
Source: GEPPAL.
Legislated quotas: Whether there was a quota law in place at the time of the last election
prior to 2009. Quota law adopted 51, no quota law 50. Source: Quota Project 2013.
Legislated aspirant quota: Whether the legislated quota concerns the aspirant stage
of the candidate selection process (party primaries). Aspirant quota 51, no aspirant
quota 50. Source: Quota Project 2014.
Level of democracy: Polity scale ranging from 110 (strongly democratic) to −10 (strongly
autocratic). Source: Polity IV Project. http://www.systemicpeace.org/polity/polity4.htm
Socio-economic development: Human Development Index (HDI) 2005. Source: UNDP.
http://hdr.undp.org/
Electoral system: Proportional representation using closed list 51; all other electoral
systems 50. Source: GEPPAL.
Sanctions (list): Whether there are sanctions for non-compliance with the required pro-
portion of women on the candidate lists. 1 5yes, 0 5no. Source: Quota project 2014.
Sanctions (rank order): Whether there are sanctions for non-compliance with rank-
order requirements of the candidates. 1 5yes, 0 5no. Source: Quota project 2014.
Appendix B National Quota Laws in Latin America (2005*), by country
Country Legislated
Quotas
Year of
legislation
Quota
provision
(lower/single
house)
Sanctions
(list)
**
Sanctions
(rank
order)
**
Argentina Yes 1991 30 Yes
[1993]
Yes [1993]
Bolivia Yes 1997 30 Yes Yes
Brazil Yes 1997 30 No No
Chile No −− − −
Colombia No
****
−− − −
Costa Rica Yes 1997 40 Yes [1999] Yes
[1999]
Dominican Rep Yes 1997 33 Yes [2000] Yes [2000]
Ecuador Yes 1997 30 No
***
No
El Salvador No
****
−− − −
Guatemala No −− − −
Honduras Yes 2000 30 No No
Mexico Yes 1996 30 Yes
[2002]
Yes [2002]
Nicaragua No
****
−− − −
Panama Yes 1997 30 No No
Paraguay Yes 1996 20 No
*****
No
Peru Yes 1997 30 No
***
No
Uruguay No
****
−− − −
Venezuela No −− − −
Notes:* GEPPAL data were collected in mid-2009. Because electoral cycles are commonly around four years
long, this table, and the variables based on it, presents the quota legislation that was in place in 2005. No new
quota laws were put in place between 2004 and 2009.
** For both sanctions for non-compliance with the required proportion of women on the candidate lists
and for non-compliance with rank-order requirements, the figures in brackets represent the year when such
rules were enforced, if not the same year as the law was passed.
*** Officially, Ecuador and Peru have sanctions (in the Peruvian case they target the proportion of women
on the lists). However, in these cases the Supreme Electoral Tribunal had not used its mandate to reject
registered lists that do not comply with the law. To the authors’knowledge, such violations of the law have
not been reported in any of the other countries. As we are concerned with strong sanctions, we consider
these countries as being without sanctions.
**** Colombia, El Salvador, Nicaragua, and Uruguay introduced quotas after the elections included in the
GEPPAL data, i.e., post-2009.
***** Paraguay has sanctions at the level of party primary elections but not at the candidate level.
Sources: Quota Project, 2013, Pacari, 2002.
Appendix C Descriptive Statistics (by party and by country)
Country Party
Wom e n
candidates (%)
Wom e n
elected (%)
No quota law
Chile Partido Radical Social Demócrata 0 0
Partido Demócrata Cristiano 10.71 10
Partido Socialista 28.57 20
Unión Demócrata Independiente 8.47 12.12
Partido Renovación Nacional 16 15.79
Partido por la Democracia 25.93 23.81
Mean (std) 14.95 (10.87) 13.62 (8.37)
Colombia Polo Democrático Alternativo 8.82 0
Partido Cambio Radical 17.39 10
Partido Liberal Colombiano 13.46 11.43
Partido Social de Unidad Nacional 14.69 13.79
Partido Conservador Colombiano 11.3 3.45
Mean (std) 13.13 (3.26) 7.73 (5.78)
El Salvador Cambio Democrático 20 0
Partido de Conciliación Nacional 27.5 0
Partido Demócrata Cristiano 30.88 20
Alianza Republicana Nacionalista (ARENA) 15.48 12.5
Frente Farabundo Martí para la Liberación
Nacional (FMLN)
36.9 31.43
Mean (std) 26.15 (8.54) 12.79 (13.48)
Guatemala Partido Unionista 17.83 0
Partido Patriota 12.82 10.34
Unidad Nacional de la Esperanza 15.67 17.65
Frente Republicano Guatemalteco 31.85 7.14
Gran Alianza Nacional 17.65 8.11
Mean (std) 19.16 (7.37) 8.65 (6.35)
Nicaragua Partido Liberal Constitucionalista (PLC) 20 8
Movimiento Renovador Sandinista 21.11 20
Frente Sandinista de Liberación
Nacional (FSLN)
32.22 31.58
Alianza Liberal Nicaragüense 36.67 9.09
Mean (std) 27.5 (8.24) 17.17 (11.03)
Uruguay Herrerismo –Partido Nacional 22.22 0
Partido Colorado –Lista 15 23.23 0
Partido Colorado –Foro Batllista 15.31 0
Asamblea Uruguay - Frente Amplio 18.57 25
Vertiente Artiguista - Frente Amplio 28.28 25
Movimiento de Participación Popular
(MPP) - Frente Amplio (FA)
16 14.81
Partido Socialista - Frente Amplio 31.31 18.18
Partido Nacional - Alianza Nacional 13.13 4.76
Mean (std) 21.01 (6.45) 10.97 (11.08)
Continued on next page.
Country Party
Wom e n
candidates (%)
Wom e n
elected (%)
Venezuela Movimiento por la Democracia
Social (PODEMOS)
17.07 0
Partido Patria Para Todos (PPT) 17.07 20
Partido Comunista de Venezuela (PCV) 17.07 25
Partido Socialista Unido 17.07 18.18
Movimiento Primero Justicia (MPJ)
(Justice First. JF)
13.13 No data
Mean (std) 16.28 (1.76) 15.80 (10.92)
Quota laws
Argentina Propuesta Republicana (PRO) - Compromiso
para el Cambio (CPC)
37.66 50
Afirmación para una República
Igualitaria (ARI)
39.33 50
Unión Cívica Radical (UCR) 40 29.41
Partido Justicialista (PJ) 37.69 35.82
Partido Socialista (PS) 43.81 36.36
Mean (std) 39.70 (2.52) 40.32 (9.25)
Bolivia Frente de Unidad Nacional (UN) 15.7 12.5
Movement Toward Socialism (MAS) 20.17 13.89
Without Fear Movement (MSM) 20.17 13.89
Poder Democrático y Social (PODEMOS) 17.32 18.6
Movimiento Nacionalista Revolucionario
(MNR)
25.51 42.86
Mean (std) 19.77 (3.74) 20.35 (12.80)
Brazil Partido Trabalhista Brasileiro (PTB) 9.7 0
Partido do Movimento Democrático
Brasileiro (PMDB)
11.47 10.11
Partido Socialista Brasileiro (PSB) 13.43 22.22
Partido Democrático Trabalhista (PDT) 10.51 4.17
Partido da Social Democracia Brasileira
(PSDB)
14.87 4.55
Partido Progressista (PP) 7.87 7.32
Democratas 12.12 7.69
Partido dos Trabalhadores (PT) 12.26 8.43
Mean (std) 11.53 (2.19) 8.06 (6.52)
Costa Rica Liberación Nacional 43.86 40
Unidad Social Cristiana 43.86 40
Movimiento Libertario 42.11 16.67
Acción Ciudadana 49.12 41.18
Mean (std) 44.74 (3.04) 34.46 (11.87)
Appendix C Continued
Continued on next page.
Country Party
Wom e n
candidates (%)
Wom e n
elected (%)
Dominican
Republic
Partido Reformista Social Cristiano No data 13.64
Fuerza Nacional Progresista No data No data
Partido Revolucionario Dominicano No data 15
Partido de la Liberación Dominicana No data 23.96
Mean (std) −17.53 (5.61)
Ecuador Partido Renovador Institucional Acción
Nacional (PRIAN)
48.28 0
Movimiento Popular Democrático (MPD) 47.32 20
Partido Sociedad Patriótica (PSP) 45.69 15.79
Movimiento País o Acuerdo País 49.14 40.68
Partido Social Cristiano (PSC) 49.25 45.45
Mean (std) 47.96 (1.48) 24.38 (18.69)
Honduras Partido Innovación y Unidad
Socialdemócrata (PINU-SD)
No data 0
Unificación Democrática 30.47 40
Partido Democracia Cristiana No data 50
Partido Nacional de Honduras No data 21.43
Partido Liberal de Honduras No data 24.19
Mean (std) 30.47 27.12 (19.14)
Mexico Partido Revolucionario Institucional (PRI) 30.2 16.5
Partido de la Revolución Democrática
(PRD)
28.37 21.26
Partido Acción Nacional (PAN) 33.2 22.82
Convergencia 28.37 26.67
Partido Verde Ecologista de México 30.2 52.63
Mean (std) 30.07 (1.98) 27.98 (14.26)
Panama Movimiento Liberal Republicano
Nacionalista (MOLIRENA)
17.5 0
Partido Político Panameñista 2.94 0
Unión Patriótica 19.51 25
Cambio Democrático (CD) 12.28 16.67
Partido Revolucionario Democrático (PRD) 14.29 7.69
Mean (std) 13.30 (6.43) 9.87 (10.89)
Paraguay Partido Patria Querida 25.64 25
Partido Democrático Progresista (PDP) 51.11 100
Partido Liberal Radical Auténtico (PLRA) 17.72 10.34
Unión Nacional de Ciudadanos Éticos
(UNACE)
31.25 13.33
Asociación Nacional Republicana (ANR) 16.25 6.67
Mean (std) 28.39 (14.08) 31.07 (39.14)
Appendix C Continued
Continued on next page.
Country Party
Wom e n
candidates (%)
Wom e n
elected (%)
Peru Partido Aprista Peruano (APRA) 36.15 22.22
Partido Popular Cristiano 38.89 29.41
Unión por el Perú 42.86 33.33
Partido Nacionalista del Perú 42.86 33.33
Cambio 90 38.89 38.46
Mean (std) 39.93 (2.90) 31.35 (6.03)
Total 24.87 (12.69) 19.17 (16.32)
Appendix C Continued
