Building Parties’ Grassroots: Electoral Systems, Party Organizations, and Social Linkages from a Cross-National Perspective ^*

Abstract

This article explores the connection between the proportional electoral system (PR) and party organizations as a key institutional determinant of party-group linkages from cross-party and national perspectives. Developing a nuanced framework, we propose an integrated model to address two questions: 01. Do candidate-centred electoral systems affect the development of party-group linkages? 02. In the case that electoral systems do have an impact, is it a direct impact or is it mediated by party organizations? Using V-Party (2020) and V-Dem (2020) databases, we selected and analyzed 617 parties in 48 countries covering third-wave democracies, post-communist countries, and the most extensive proportional democracies in Europe (Western and Eastern Europe) and Latin America between 1989 and 2019 – Large- N cross-national comparative analysis (JANDA, 1980). Based on panel models, we found that the candidate-centred electoral system is negatively related to the development of strong ties between parties and groups, but only in the case of party organizations with low levels of party strength, intraparty cohesion, and financial linkages with non-party groups. When decision-making powers are concentrated in the hands of powerful party elites, these elites can solve coordination problems, mitigate intraparty conflicts, and deal with the consequences of personalization. Therefore, intraparty politics varies empirically since parties respond to their contextual challenges (electoral rules) strategically, with consequences for party-group linkages.

Party organizations; comparative politics; party change; party-group linkages; electoral systems

This article explores the connection between proportional (PR) electoral system and intraparty organizations as a key institutional determinant of party-group linkages from cross-party and cross-national perspectives. A considerable body of literature on electoral systems emphasizes that candidate-centred electoral systems – which encourage candidates to cultivate personal instead of party reputations – promote intraparty competition and rootless party organizations ( AMES, 2001 ; ANDRÉ et al., 2014 ; CAREY and SHUGART, 1995 ; CHANG and GOLDEN 2007 ; COLOMER, 2011 ; MAINWARING, 1999 ; SHUGART, 2005 , 2001 ; PASSARELLI, 2020 ).

Nonetheless, there is little comparative empirical research on how party organizations have strategically responded (or adapted) to changes in electoral rules that foster institutional personalization ( BALMAS et al., 2014 , pp. 38-39)² . Notably, most research has an omitted variable bias as party-level dimensions are excluded, particularly organizational aspects of political parties ( PASSARELLI, 2020 , p. 21). We address two questions: 01. Do candidate-centred proportional electoral systems affect the development of party-group linkages? 02. In the case that the proportional electoral system does have an impact, is it a direct impact, or is it mediated by intraparty organizational structures?

Based on longitudinal and multilevel approaches, we expect that candidate-centred electoral systems would be negatively related to the development of strong ties between parties and society, but only in the case of party organizations with weak structures and high levels of inclusiveness in intraparty decision-making processes. When parties have a robust organization, and when the decision-making power is concentrated in the hands of a few, party elites can enforce their decisions, solve coordination problems, and mitigate personalistic incentives in electoral systems – they preserve the collective interests of political parties to the detriment of exclusively individualistic ambitions ( ALDRICH, 2011 ; RAHAT and KENIG, 2018 ).

Therefore, we argue that structures, strategies, and decisions at the party-level play a mediating role in how the electoral system affects party-group linkages ( ALLERN et al., 2007 ; ALLERN and VERGE, 2017 ; THOMAS, 2001 ). Our article offers a roadmap on the interaction between incentives from electoral rules and party organizations, emphasizing their consequences for those linkages. Changes in electoral systems elucidate what the parties’ electoral tactics are and how power relations are structured within them ( COLOMER, 2011 ; SHUGART 2005 ). However, party candidates do not compete in an organizational vacuum: Our analysis considers the influence of intraparty dynamics – particularly party organizations – in which candidates operate.

From a theoretical perspective, different electoral systems offer distinct incentives for candidates, party members, and voters, structuring representational linkages between citizens and representatives ( CAREY and SHUGART, 1995 ; SHUGART, 2005 , 2001 ). This study focuses exclusively on proportional (or semi proportional) systems (PR), mainly because by doing so it is possible to clearly distinguish electoral systems in which voters must (or prefer to) cast a vote (or votes) for a party from those in which voters can rank parties and/or candidates.

Following the tradition of the 'most different' approach in a comparative analysis of party organizations ( JANDA, 1980 )³ , we rely on the two most extensive cross-national datasets on party organizations and elections, which are both based on expert evaluations: V-Party Dataset ( LÜHRMANN et al., 2020 ) and V-Dem Dataset ( COPPEDGE et al., 2020 )⁴ . Our conclusions offer a more nuanced picture than that found in previous research; most importantly, our results suggest that the scholarly debate should ascribe a more central role to electoral and organizational dimensions, acknowledging that these dimensions may have a strong influence on substantive political outcomes, particularly on party-group linkages.

We propose specific empirical measures based on party literature to capture variations in intraparty decision-making and organizational capacity dimensions: 01. party strength, 02. financial linkage (financial support from party members and civil society), 03. inclusiveness in candidate selection processes, and 04. intraparty cohesion. However, comparative research on intraparty politics remains limited, mainly because of the lack of reliable data ( BORZ and JANDA, 2018 ; SHUGART, 2005 , p. 36). To close this gap, Party V-Dem provides a longitudinal party-level dataset that allows researchers to access the impact of internal party dynamics on different democratic outcomes ( DÜPONT et al., 2022 , 2021 )⁵ .

We analyzed 617 parties from 48 democracies that adopt a proportional (PR) electoral system (including SNTV and STV) to verify the impacts of electoral systems and party organizations on party-group linkages. To enhance comparability, the dataset covers third-wave democracies, post-communist countries, and most democracies in Latin America and Europe (Western and Eastern Europe) in the period between 1989 and 2019 (see Table A02 – Appendix). All parties in the dataset won at least 05% of the votes in each legislative election. Following the Freedom House parameters, we did not add authoritarian or semi-authoritarian regimes ('not free countries')⁶ . Being related to the V-Dem dataset, the V-Party⁷ is based on expert evaluations and follows the same methodology⁸ . However, while countries in the V-Dem dataset are updated every year, political parties are only codified when they run for national elections.

Although widely employed in the comparative literature, expert-oriented data generation carries some potential limitations, particularly in the case of the V-Dem dataset and, by extension, the V-Party dataset. By and large, at least four coders have provided their assessment per observation to guarantee the reliability and conformability of the datasets ( DÜPONT et al., 2022 , 2021 ; LÜHRMANN et al., 2020 ). Even so, experts may have divergent opinions or make mistakes during the codification process.

To address this challenge, both projects (V-Dem and V-Party) adopted the Bayesian Item Response Theory (IRT), which aggregates expert ordinal ratings into point estimates of latent (or unobservable) concepts and quantifies the uncertainty around the interval estimations, considering the pattern of distinction, connections, and parallel coding ratings ( COPPEDGE et al., 2020 ; DÜPONT et al., 2021 ; LÜHRMANN et al., 2020 ; PEMSTEIN et al., 2020 ). After performing several validation tests with multiple previous datasets on political parties, Düpont et al. (2021) concluded that "party data aligns well with extant data on party organizational characteristics both from more recent surveys as well as older data reaching back to the 1980s. This puts confidence in utilizing V-Party data for longitudinal and cross-country analyses" ( DÜPONT et al., 2021, p , p. 17 (see also DÜPONT et al., 2022, p , p. 06)).

Thus, despite potential limitations, V-Dem and V-Party databases provide us with an unprecedented opportunity to explore the consequences of changes in institutions and parties' structures over a long period. Moreover, combining V-Dem and V-Party datasets allows researchers to access the impact of electoral institutions and internal party dynamics on party-group linkages through a robust large-N cross-national comparative design ( JANDA, 1980 ). As noted by Benoit and Laver (2006): "the great virtue of an expert survey is that it sets out to summarize the judgments of the consensus of experts on the matters at issue, and moreover to do so in a systematic way (…) expert survey results provide a benchmark that gives some systematic sense of the content validity of alternative measures" ( BENOIT and LAVER, 2006 , p, p. 03).

The remainder of this paper is organized as follows. First, we summarize the theories of electoral arrangements and party organizations and hypothesize their potential impacts on party-group linkages that we examine empirically. Next, we present and describe a cross-national empirical dependent variable and independent variables. Then, empirical models and results are discussed. The last section concludes.

Theoretical debate and hypothesis

V.O. Key Jr. (1964) introduced the concept of ‘linkage’ to designate the capacity of parties to establish connections between citizens (mass opinion) and governments ( KITSCHELT, 2000 ; LAWSON and MERKL, 1988 ). Comparative literature has shown that the linkages between political parties and society have declined in many democracies, mainly in those that are most consolidated. This general trend of decline, however, has not been homogeneous or linear over time and across regions; some parties have experienced periods of growth in their capacity to connect ( RAHAT and KENIG, 2018 ).

Party literature has primarily highlighted the modernization approach: changes in lifestyle, citizens' increasing distrust in political parties, and modern campaign techniques. According to this approach, although party organizations have preserved some procedural functions in government and in the state apparatus ( BARTOLINI and MAIR, 2001 ; IGNAZI, 2012) , they came to be perceived as antiquated or even irrelevant in their relationships with citizens ( DALTON and WATTENBERG, 2000 ; LAWSON and MERKL, 1988 ; MAIR, 2013 ; SCHMITTER, 2001 ; WATTENBERG, 1994) .

There is a recent renewed interest in how political parties still mobilize their organizational structures to establish strategic channels with social groups and organized citizens. From this perspective, the ‘collective terms of agreement’ are highlighted to explain the cost-benefit interactions between power-seeking elites and organized interests ( ALLERN et al., 2020 ; ALLERN et al., 2007 ; ALLERN and BALE, 2012 ; ALLERN and VERGE, 2017 ; BAWN et al., 2012 ; POGUNTKE 2006 , 2002 ; THOMAS, 2001 ). The emphasis now is on examining the relationship between political parties and social groups through a non-zero-sum game perspective. In more conceptual terms, according to Poguntke (2002): "organizational linkage facilitates two-way communication between party elites and groups of voters, mediated through organizational channels and based on the exchange of electoral mobilization for policy responsiveness” ( POGUNTKE, 2002 , p. 07).

In this sense, party elites are gatekeepers of social-group influences. In other words, the elites strategically invite social groups to establish relations with intraparty structures, bringing attention to their advantages or attacking other parties and organized interests. More specifically, they are agenda-setters – not only reactive but also proactive – in their relationships with different social groups ( PRZEWORSKI and SPRAGUE, 1986 ; SARTORI, 2005 ; THOMAS, 2001 ).

In addition to encouraging party-system institutionalization ( MAINWARING, 1999 , p. 325), party-group linkages – if systemic and relevant – create a sense of democratic legitimacy for political parties by enhancing their representative functions ( IGNAZI, 2012 ; MAIR, 2013 ). In short, on the one hand, some scholars have questioned the parties’ capacity to develop ties with social groups, particularly in the case of established parties – Social Democratic, Conservative, and Christian Democrat parties ( MAIR, 2013 , pp. 37-38). On the other hand, empirical evidence also suggests that some parties should be more concerned about maintaining their social linkages: trajectories of party-group linkages vary considerably from case to case and they may remain stable over time ( ALLERN et al., 2007 ; ALLERN and BALE, 2012 ; POGUNTKE, 2006 , 2002 ; THOMAS, 2001 ).

In general, the discussion thus far has highlighted some external determinants driven by social changes in party-group linkages, but relatively little has been said about the centrality of the institutional environment⁹ . We intend to address this gap by exploring party structures as intermediate variables between electoral systems and party-group linkages, emphasizing organization-based relationships.

Fundamentally, we assume here that electoral systems are structural elements, or "unstable constants", so the key aspect is how electoral systems affect intraparty conflicts and linkage strategies over time. Carey and Shugart's seminal article (1995) sorts the incentives of electoral systems into two polar categories: candidate-centred and party-centred electoral systems. From a 'top-down' (or elite) perspective, when the 'preferential vote' is allowed, politicians have more incentives to cultivate personal reputations than party reputations ( ANDRÉ et al., 2014 ; CAREY and SHUGART, 1995 , p. 419; COLOMER, 2011 ; DALTON et al., 2011 ; PASSARELLI, 2020 ; RENWICK and PILET, 2016 ; SHUGART, 2005 , 2001 ). Moreover, the authors emphasize that a personalized electoral system could affect candidates’ and legislators' behavior (the 'style') ( MAYHEW, 1974 ), thus resulting in political instability in new democracies ( AMES, 2001 ).

In electoral systems that favor institutional personalization, the political relevance of individuals increases, while political parties – as a collective enterprise – decline ( RAHAT and KENIG, 2018 ). The usual expectation is that candidate-centred electoral systems would incentivize intraparty conflicts and encourage candidates to distinguish themselves from their parties and co-partisans. Therefore, a linkage strategy that focuses on directly contacting individual politicians is more effective for non-party groups.

As André, Depauw, and Shugart (2014) argue: "electoral institutions, in particular, structure the options available to voters and thereby strengthen or weaken the incentives for legislators to cultivate their personal reputations rather than the party's reputation” ( ANDRÉ, DEPAUW, and SHUGART, 2014 , pp. 231-232). It should thus be expected that the linkage between parties (as an organized group) and society would be undeveloped in personalized electoral systems.

Hypothesis 01: candidate-centred electoral systems are negatively related to the development of strong ties between parties and groups due to intraparty competition and because party-group linkages are less valuable electoral instruments for communicating with constituencies. Hence, the power balance between party-group linkages, on the one hand, and individualistic power-seeking strategies, on the other, are related to the electoral system’s degree of personalization.

Is the impact of electoral systems on expected party-group linkages mediated by party organizational characteristics? From the perspective of intraparty politics, electoral rules constrain organizational strategies, but they do not determine the choices of party actors ( ALLERN and BALE, 2012 , p. 17; HARMEL and JANDA, 1994 ). Thus, as shown in Figure 01 , we expect that party organizational characteristics would mediate institutional incentives for establishing linkage strategies. As Allern and Verge point out (2017): "the party level is crucial for how parties choose to organize their relationships with social groups. Parties are autonomous, power-seeking actors that choose to organize in ways that help them efficiently achieve their goals” ( ALLERN and VERGE, 2017 , p. 110).

Figure 01
Summary of the argument: a multilevel model

We argue that different party organizations may amplify, obstruct, or offset the effects of electoral institutions on party-group linkages. Party elites are not always passive in the face of electoral incentives ( COX, 1997 ): They anticipate (or react to) the effects of these incentives and mobilize organizational resources to limit intraparty competition and enhance parties as a collective enterprise. As a result, organizational mechanisms can be mobilized to mitigate rebellions (and factions), sustain party-oriented careers, and develop party reputation.

Thus, by controlling these mechanisms, party elites can make more credible agreements and promises (or threats) to party members and social groups. In addition, these organizational resources can help develop more predictable electoral performance and party policy positions. From the perspective of non-party groups, making agreements with robust and stable parties is more advantageous than doing so with transitory individuals ( ALESINA and SPEAR, 1988 ).

Hypothesis 02: When party elites have the control of a robust and cohesive party organization – one that also includes a strong financial support from society and a low level of inclusiveness in candidate selection processes –, the most robust forms of party-group linkages are stimulated. Conversely, when political parties are organizationally weak, non-cohesive, and porous to personal interests, their members will act as independent entrepreneurs seeking to establish connections with social groups rather than act as team players.

The multiple pathways of party-group linkages

Focusing on variations in party-society linkages, the first step is to identify recent party-group relationships. Our proposal for operationalization thus combines the different perspectives that are prevalent in the literature. On the 'demand side,' the emphasis lies in the structured ties (formal or informal) that party operators (elites) encourage or mobilize in their relationship with civil society organizations, collateral organizations, social movements, and trade unions ( ALLERN et al., 2020 ; ALLERN and VERGE, 2017 ; POGUNTKE, 2006 , 2002 ). On the 'supply side,' the focus is on how socially consistent party organizations are, that is the extent to which sociodemographic characteristics – and social identity – of active supporters are congruent (or aligned) with parties’ electoral basis ( BARTOLINI, 2000, p , pp. 26-27; BAWN et al., 2012 ), thus covering various social sectors, from aristocrats and the military to working-class people, women, large businesses, and religious groups.

In other words, while one perspective allows us to better assess whether parties spend considerable effort to build organizational ties with society ( SARTORI, 2005 ), the other addresses the ways in which social heterogeneity plays out within parties' organizational structure. These two perspectives are conceptually complementary: a close relationship with organized groups may help mobilize latent core supporters in and between elections – as previously mentioned, these aspects encompass formal and informal strategies, including 'new' (e.g., postmaterialist movements or subunits) and 'traditional' (e.g., trade unions) groups.

To outline a comprehensive scenario for proportional systems (PR), and bearing in mind the predominant conceptualizations of party-group linkages used in comparative literature, we standardized (z-scores) and added two V-Party variables: v2pagroup¹⁰ and v2pasoctie¹¹ (see Figure A01 - Appendix ) . Hence, the dependent variable is constructed according to following expression: ‘party-group linkage index = affiliate organizations’ (v2pagroup) + ‘ party support group’ (v2pasoctie)¹² . In a recent article, a similar empirical strategy was used by Düpont et al. (2022 , p. 02). Full details about questions and codification schema are available in Chart A01 - Appendix .

Figure 02 presents an overview of average trends in party-group linkages in all 48 democracies covered by our sample over time. Despite the diagnosis that linkages between political parties and society have declined in new and traditional democracies ( DALTON and WATTENBERG, 2000 ; MAIR, 2013 )¹³ , there is no general declining trend. Instead, we have identified, on average, a pattern of relative stability in recent decades. Figure 03 shows distinctive tendencies for party-group linkages when segregated into three regions. In Eastern Europe, such linkages have increased progressively until 2010, followed by a relative decline. In the same period, Latin American countries experienced stabilization, preceded by a destabilization cycle. Consistent with the diagnosis of party decline, Western Europe had the most pronounced retraction in social linkages until 2010. Subsequently, Western European parties partially recovered their capacity to establish a relationship with organized society. Contrary to what is expected by modernization theories, our initial findings suggest that there is no obvious evolutive pattern of party-group linkages across regions¹⁴ .

Figure 02
Party-group linkage index over time (on average)

Figure 03
Party-group linkages across regions (on average)

Consequences of political parties and electoral rules beyond the election day: party organizations and electoral systems as independent variables

Measuring electoral systems: life under party lists

Electoral rules have consistently been identified as a potential key factor in shaping party decisions ( COX, 1997 , p. 38). Electoral systems in particular generate political opportunities and incentive structures for incumbents or ambitious politicians to build and perpetuate their party organizations ( ALDRICH, 2011 ; STROM, 1990 ). To operationalize more or less permissive electoral rules, we adopted a synthetic and modified Shugart's index idealized by David Farrell and Roger Scully ( FARRELL and McALLISTER, 2006 ; FARRELL and SCULLY, 2007, p , pp. 128-129; SHUGART, 2001 ) — the 'modified Intraparty efficiency index.' According to Farrell and Scully (2007, p , p. 128), this index synthesizes three main characteristics of electoral systems (ballot, vote, and district) into a unique single measure ( SHUGART, 2001, p , p. 182). Moreover, the 'modified intraparty efficiency index' allows us to detect the extent to which electoral success depends on either nominal or list votes ( FARRELL and McALLISTER, 2006 ; FARRELL and SCULLY, 2007 ; SHUGART, 2001, p , p. 182).

The codification process is summarized in Chart 01 ; all scores in our set of 48 nations are in Figure A01 - Appendix . We also consider the average district size as a separate variable (v2elloeldm)¹⁵ . As previously mentioned, our interest here is in systematic developments and their consequences over time¹⁶ . Higher scores across these components suggest a candidate-centred electoral system (maximum score: 09 – SNTV), while lower scores indicate a party-centred electoral system (minimum score: 03 – closed list). Data on electoral systems cover 48 countries and are derived from the V-Dem project (2020) and from supplementary sources. With this coding strategy, we can locate all democracies and their transformations over time on a scale from predominantly candidate-centred to predominantly party-centred.

Chart 01
Modified intraparty efficiency index

	Coding	Theoretical rationale
Ballot	1 = Ballot access controlled by parties, and voters may not disturb the order of the list. 2 = Ballot access dominated by parties, but voters may disturb list. 3 = Ballot access nearly unrestricted.	The more control the party has over the ballot, the greater is the potential incentive for candidates to develop a personal reputation
Vote	1 = Vote for list only. 2 = Vote is list or nominal, but list votes predominate. 3= Vote is nominal or list, but nominal votes predominate and pool to other candidates. 4 = Vote is nominal only, but the vote may pool or transfer to other candidates.	Preferential vote decreases the relevance of party reputation.
Magnitude	1 = District magnitude is greater than one, with Vote < 3. 2 = District magnitude is greater than one, with Vote > 2, provided that Ballot > 1.	In party-centred electoral systems, the higher the district magnitude, the lower the relevance of candidate reputation. Meanwhile, in candidate-centred electoral systems, the higher the district magnitude, the higher the relevance of candidate reputation.
Aggregation	$\text{Score index}=\text{Ballot}+\text{Vote}+\text{Magnitude}$

The final dataset includes 19 Western European countries, 15 Latin American countries, and 14 Eastern European post-communist countries. Including electoral reforms in the period, we analyzed 25 closed-list PR, 21 open-list PR (including SNTV and STV), and 11 flexible systems from 1989 to 2019.

Figure 04 displays the average for standardized scores over time (z-scores). We aggregated averages to simplify the presentation – more detailed scores by country are available in Figure A01 – Appendix (min: -1.07, max: 2.1). By using five-year intervals, we minimized short-term average variation to focus on long-term trends. Overall, there are only modest and nonlinear net changes in the direction of candidate-centred electoral systems. The overall change from 1989 to 2019s – across all democracies – was nearly 0.03.

Figure 04
Modified intraparty efficiency' index over time (on average)

Some countries, however, show significant temporal variation in distinct directions. For instance, some electoral reforms enhance candidate-centred aspects of electoral systems, such as those adopted in Chile, the Dominican Republic, and Iceland, while others go in a different direction, such as the Italian and Colombian reforms¹⁷ . These findings are consistent with those of Rahat and Kenig (2018 , p. 142) and Renwick and Pilet (2016) in European countries, Australia, Japan, New Zealand, Israel, and Canada. Hence, there is considerable variation between regions. Table 01 also demonstrates that Western Europe has on average more personalized proportional (PR) electoral systems than Eastern Europe and Latin America (sig < 0.05).

Table 01
Independent variables across regions and over time

	Internal cohesion		Inclusiveness in candidate selection			Party strength		Financial linkage		The modified intraparty efficiency index		Magnitude average (M)
Years
	Mean (Std. Dev)	N		Mean (Std. Dev)	N	Mean (Std. Dev)	N	Mean (Std. Dev)	N	Mean (Std. Dev)	N	Mean (Std. Dev)	N
1995	0.3 (1.2)	418		0.43 (1.3)	429	-0.17 (1.8)	429	0.24 (1.5)	429	0.06 (1.0)	42	17.8 (33)	42
2000	0.2 (1.2)	255		0.44 (1.3)	260	0.03 (1.9)	260	0.08 (1.4)	259	-0.92 (0.99)	42	34.7 (63)	42
2005	0.23 (1.2)	251		0.62 (1.3)	256	0.06 (1.8)	256	0.07 (1.4)	255	0.90 (1.0)	41	22.7 (40)	41
2010	0.20 (1.2)	347		0.51 (1.4)	352	0.06 (1.7)	352	-0.25 (1.4)	350	-0.02 (0.92)	45	44.1 (83)	45
2015	0.16 (1.3)	332		0.47 (1.5)	344	0.06 (1.7)	340	-0.06 (1.5)	342	0.02 (1.0)	44	29.8 (52)	44
2019	0.17 (1.2)	283		0.57 (1.5)	289	0.05 (1.7)	288	-0.11 (1.4)	285	0.09 (0.97)	42	29.9 (51)	42
Eta-squared	0.003			0.002		0.003		0.013***		0.004		0.024***
Regions
Latin America	0.07 (1.18)	602		0.05 (1.4)	602	-0.70 (1.7)	602	-0.14 (1.1)	602	-0.11 (1.05)	15	6.6 (4.9)	15
Eastern Europe	0.04 (1.2)	537		0.011 (1.2)	537	0.29 (1.7)	537	-0.29 (1.0)	537	0,16 (0,9)	14	56.8 (85.8)	14
Western Europe	0.51 (1.3)	748		1.17 (1.2)	791	0.35 (1.7)	790	0.31 (1.9)	781	0.14 (0.98)	19	28.1 (44.3)	19
Eta-squared	0.031***			0.157***		0.068***		0.032***		0.011***		0.114***

• Source: V-Party (2020) and V-Dem (2020).
• Notes: n = 1930. Eta squared refers to the between-groups variance explained by regions and years, *** p<0.01, ** p<0.05, * p<0.1.

Party organizations: a multidimensional phenomenon

With respect to how electoral dynamics affect party-group linkages, we ask: Do party organizations make a difference? Parties' characteristics widely differ across regions, countries, and within the same party system, potentially with significant political implications ( JANDA, 1980 ; POGUNTKE et al., 2016) . In this sense, the same electoral institutions may be associated with quite different intraparty outcomes ( CHHIBBER et al., 2014) .

Given our broad organizational focus here, we do not presume that all intraparty spheres evolve symmetrically in the same environments, also known as institutional isomorphism ( OLIVER, 1991 ). As Barnea and Rahat (2007) emphasized: “at the end of the day decisions are taken internally, with a certain level of autonomy for the decision-makers” ( BARNEA and RAHAT, 2007 , p. 377). Based on previous comparative research, we present organizational dimensions that are related to central variables in party literature: 01. party strength, 02. financial linkage (financial support from party members and civil society), 03. inclusiveness in candidate selection processes, and 04. intraparty cohesion ( HAZAN and RAHAT, 2010 ; JANDA, 1980 ; KATZ and MAIR, 1995 ; TAVITS, 2013 , pp. 17-18).

Moreover, these variables accommodate a wide range of potential factors that affect party-group linkages, and they also describe aspects of party structures entirely or mainly under the control of party elites ( HARMEL and JANDA, 1994 , p. 261). Specifically, we emphasize that preferences, priorities, and intraparty decision-making may affect the linkage strategies of party elites. This argument implies that rational leadership decisions can establish or abolish those aspects ( ALDRICH, 2011 ; HARMEL and JANDA, 1994 ). The existing literature on political organizations provides the basis for four central variables:

The first variable, ‘party strength’ , consists of two conceptually distinct analytical dimensions: 01. party activists and permanently active personnel (bureaucracy) in local communities; 02. permanent offices at the local level. At the party level, organizational strength comprises mobilization capacity, professionalization, and parties’ territorial extension beyond parliamentary structures ( BIZZARRO et al., 2018 ; COLEMAN, 1996 ; COTTER et al., 1989 ; RIBEIRO and LOCATELLI, 2019 ; TAVITS, 2013 )¹⁸ . Tavits (2013) argues that "party has a strong organization if it has structures, personnel, and activities beyond public office” ( TAVITS, 2013 , p. 19).

In short, we expect a positive relationship between party strength and party-group linkages ( THOMAS 2001 , p. 279): Strong parties are more effective in reaching close and long-term attachments with organized groups, establishing face-to-face contact with citizens, and recruiting social leadership ( TAVITS, 2013 ). Moreover, from the social group's perspective, parties with robust organizations are more electorally viable and more trustworthy when it comes to establishing connections. The party strength score is a composite variable created by adding the standardized (z-score) local organizational strength (v2paactcom)¹⁹ and local party office (v2palocoff)²⁰ for each party. Not surprisingly, as Table 01 shows, Western European parties are more robust than those in Eastern Europe and Latin America (sig. < 0.05). Conversely, on average, there is no significant difference in the evolution of party strength over time (min: -4.7; max: 3.8).

The second variable, ‘ financial linkage’²¹ , refers to party funding from party members and organized groups. One of the most important aspects of party organizations concerns how they finance themselves. In some parties, fee collection (to parties directly or indirectly) is central to ensuring intraparty accountability between party elites, mass membership, and organized interests ( KATZ and MAIR, 1995 ; MAIR, 2013 ). When party leaders are dependent on group resources, they increase their investment in linkages ( THOMAS, 2001 ). Contributions from party members and civil society groups provide political parties with not only monetary benefits but also legitimacy ( IGNAZI, 2012 , pp. 75-76).

An additive and standardized index (z-score) was created to access relevant parties’ financial connections with non-party organizations and party members. To this end, we aggregated the scores of parties that have used – as a primary financial source – the donations from civil society organizations and trade unions (v2pafunds_3) and party membership fees and supporters’ small-scale donations (v2pafunds_4)²² . As shown in Table 01 , and consistent with findings in the literature, our data suggest that financial support from both party members and non-party organizations has significantly decreased over time ( KATZ and MAIR, 1995 ; MAIR, 2013 ). Additionally, Table 01 reports results that are significantly different across regions; the financial linkage average in Latin America and Eastern Europe is lower than that in Western Europe (min: -1.5, max: 5.3, sig. < 0.05).

The third variable, ‘candidate selection’, refers to the important process through which political parties nominate their candidates for electoral disputes ( HAZAN and RAHAT, 2010 ). Our focus is on the size of the selectorate who can participate in the selection procedure and on its consequences – inclusiveness. We expect that exclusive candidate selection processes favors intraparty coordination, limits intraparty competition, and generates a more representative candidate list ( HAZAN and RAHAT 2010 , p. 114). To some extent, control over ballot access is a critical currency in bargaining processes; this, in turn, stimulates the establishment of party-group linkages.

In addition, if party elites, as gatekeepers, control re-nomination, they have the tools to discourage intraparty conflicts (factionalism and personalism) in the face of contextual challenges ( COLOMER, 2011 ; HAZAN and RAHAT, 2010 ; KEY 1964 ). Using the standardized candidate nomination (v2panom)²³ , the higher the index score, the greater the inclusiveness in candidate selection processes. As the second column in Table 01 indicates, candidate selection varies significantly across regions (min: -2.9, max: 4.6, sig. < 0.05).

Finally, ‘intraparty cohesion’ relies on party members sharing the same preferences — particularly, we expect that the more robust party cohesion is, the stronger party-group linkages will be. In these terms, a cohesive behavior may sustain the party’s reliability and accountability with non-party organizations, image. When combined, all these factors make parties significantly predictable: an incoherent behavior from party elites emphasizes short-term strategies such as resorting to appealing candidates, media-based campaigns, and populist (or flip-flop) policies to attract group support ( ALDRICH, 2011 ; JANDA, 1980 , p. 118; TAVITS, 2013 ).

As for the variable labeled 'internal cohesion' (v2padisa)²⁴ , a positive change increases intraparty cohesion. Table 01 demonstrates that the average party-group linkages is lowest in Eastern Europe and Latin America, followed by Western Europe (min: -4.1, max: 2.9, sig. < 0.05).

Overall, we recognize the role of institutional frameworks (contextual factors) in shaping parties' behavior; we also acknowledge that actors have some leeway in how they use the opportunities offered by the institutions ( BARNEA and RAHAT, 2007 ; CHHIBBER et al., 2014) .Therefore, in examining electoral-system and party-level determinants, we must also control multiple external elements – control variables are as follows (summary statistics are available in Table A01 - Appendix ):

Institutional controls

Presidentialism Index (v2xnp_pres): presidentialism can affect party organizations and party-group linkages by enhancing intraparty conflicts between parliamentarians and presidential candidates ( THOMAS, 2001 ). Particularly, each of these groups caters to different electoral constituencies (local and national) – competing principals' dilemma from a principal-agent perspective ( KEY, 1964 ; MAINWARING, 1999 ; SAMUELS and SHUGART, 2010) . Thus, it is expected that highly presidentialized governments would have very few party-group linkages – Interval, from low to highly presidentilized government (0-1).

Division of power index (v2x_feduni): the division of power in federal and unitary democratic regimes may undermine party-group linkages ( THOMAS, 2001 ). In a less cooperative institutional framework, parties are more autonomous and cohesive. Hence, federalism weakens parties because it reinforces existing sub-national conflicts and generates new sources of conflict ( AMES 2001 ; KEY, 1964 ; MAINWARING 1999 ). Moreover, where power is decentralized, state and federal parties tend to be hierarchically and ideologically less integrated ( AMES, 2001 ). Both factors represent challenges to building stable linkages between parties and organized groups. Therefore, the greater the positive variation in the division of power index, the greater the autonomy of subnational governments – scale (0-1).

Compulsory voting (v2elcomvot): we argue that compulsory voting laws enforced via costly sanctions can reduce the incentives for political parties to build linkages with social groups. Parties are not much dependent on mobilization, which decreases the incentives for leaders to invest in linkage strategies ( DALTON et al., 2011 ). Thus, compulsory voting reduces the electoral relevance of establishing party organizations and, by implication, makes party-group linkages less important. A positive variation in the indicator means a higher level of sanctions on compulsory voting – scale (0-3).

Socio-political controls

Civil society index (v2xcs_ccsi): we expect that the growing distance between parties and society would derive from the fact that alternative forms of participation are increasingly available in a more mobilized civil society. Such a context may produce systemic consequences for party-social group linkages, making it more difficult for parties to perform their representative and aggregative functions ( DALTON and WATTENBERG, 2000 ; LAWSON and MERKL, 1988 ; SCHMITTER, 2001 ). Therefore, a positive variation in the indicator would mean higher civil society mobilization – scale (0-1).

Educational equality (v2peedueq): scholars assert that higher levels of economic development are associated with a more broad-based distribution of educational and occupational resources. Greater access to educational and occupational resources increases engagement opportunities on alternative (or new) forms of participation in the face of party decline. Therefore, we expect that higher levels of educational equality would lead to fewer party-group linkages ( DALTON and WATTENBERG, 2000 ; MAIR, 2013 ; SCHMITTER, 2001 ) – a positive variation in the indicator would mean educational equality – scale (0-4).

Results and discussion

Bivariate analysis

Party literature is interested in how institutional choices affect democratic effectiveness; it suggests that certain electoral systems are likely to encourage (or not) strong party organizations ( AMES, 2001 ; MAINWARING, 1999 ; PASSARELLI, 2020 ). Table 02 presents the Pearson correlations between all our indicators for party organizations and electoral systems. On the one hand, the 'modified intraparty efficiency index' correlates positively and significantly with financial linkage, inclusiveness in candidate selection processes, and internal party cohesion (sig. <0.05). On the other hand, the correlation between party strength is negative and not significant. In essence, the analysis suggests that there is no perfect correlation between the electoral system characteristics and intraparty dimensions, notably organizational variables since correlation levels are weak.

Table 02
Pairwise correlations between 'modified intraparty efficiency index' and party strength, internal cohesion, financial linkage, and candidate selection (inclusiveness)

Variables	Party strength	Financial linkage	Candidate selection	Internal cohesion
Financial linkage	0.061***
inclusiveness in candidate selection	0.265***	0.116***
Internal cohesion	-0.070***	0.020	-0.053**
Magnitude average	-0.116***	-0.087***	-0.142***	-0.014
Modified intraparty efficiency index	-0.008	0.302***	0.103***	0.119***

• Source: V-Party (2020) and V-Dem (2020).
• Notes: n = 1930; *** p<0.01, ** p<0.05, * p<0.

Moreover, in general, there is a weak negative correlation between the average district magnitude and the party organizational dimensions (except for party strength). Consequently, when electoral magnitude is high, parties tend to limit their levels of inclusiveness in candidate selection processes, party strength, and financial linkages (sig<0.05). Hence, district magnitude is associated with distinct party organizational characteristics – parties tend to be less inclusive and more limited in their organizational capacity. Last but not least, as indicated in Table 02 , party organization variables are not strongly correlated, which justifies analyzing them as distinct dimensions. In this sense, parties can develop internal procedures and structures following non-symmetrical orientations ( RAHAT and KENIG, 2018 ; TAVITS, 2013) .

Multivariate analysis

To evaluate the extent to which electoral systems and party organizational changes increased the intensity of party-group linkages, Table 03 presents estimates of six general models of panels with fixed effects, utilizing all parties that won at least 05% of the votes between 1989 and 2019. Using the entire sample, we estimate panel models with fixed-effect to analyze the changes within a dimension – party-group linkages – and control for all time-invariant differences²⁵ . So, we controlled for unobserved heterogeneity by adding the fixed-effects at regions, countries, and party levels, thus estimating the effects of 'modified intraparty efficiency index', average district magnitude, and party organization variables on party-group linkages ( WOOLDRIDGE, 2002 ).

Table 03
Results for panel analyses with fixed effects

Variables	(1)	(2)	(3)	(4)
	Electoral system	Party organizations	Electoral-organizational	Complete model
Electoral system
District magnitude (M)	-0.011		-0.009	-0.008
	(0.009)		(0.008)	(0.008)
Modified intraparty efficiency index	-0.094***		-0.061**	-0.062**
Party Organization	(0.031)		(0.030)	(0.030)
Party strength		0.438***	0.435***	0.425***
		(0.030)	(0.030)	(0.030)
Financial linkages		0.068***	0.061**	0.044*
		(0.026)	(0.026)	(0.026)
Internal cohesion		0.055***	0.053***	0.050***
		(0.013)	(0.013)	(0.013)
Inclusiveness in candidate selection process		-0.025	-0.013	0.046
Institutional variables		(0.028)	(0.028)	(0.029)
Presidentialism index				0.357
				(0.224)
Division of power index (federal-unitary)				-0.008
				(0.138)
Compulsory voting				0.108*
Socio-political variables				(0.056)
Civil society index				-0.132
				(0.168)
Educational equality				0.040
				(0.038)
Constant	0.317	-0.037***	0.223	0.124
	(0.259)	(0.017)	(0.251)	(0.308)
Observations	1,922	1,868	1,868	1,868
Within R-squared	0.009	0.150	0.154	0.164
Number of parties	611	600	600	600
Parties FE	YES	YES	YES	YES
Countries FE	YES	YES	YES	YES
Regions FE	YES	YES	YES	YES

• Source: Created by the author, based on V-Party (2020) and V-Dem (2020) datasets.
• Notes: Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1.; dependent variable: party-group linkage index.

Table 03 presents different variables (organizational, socio-political, and institutional) structured in four parts. Model 01 and 02 includes, respectively, variables related to the electoral systems and party organizations – additionally, Model 03 aggregates these two sets of variables, and Model 04 also include all control variables.

The 'modified intraparty efficiency index' is negatively and significantly related to party-group linkages: the more the PR electoral systems are candidate-centred, the fewer linkages a party is likely to develop. However, as expected (hypothesis 01), the 'modified intraparty efficiency index' steadily and consistently affects party-group linkages in all four models. Moreover, in adding different variables, the intensity of the relationship remained relatively steady. As Model 01 reveals, by adding one point to the modified intraparty efficiency index' (min: -1.07, max: 2.1), a reduction of approximately -0.09 points by year (min: -3.4, max: 6.0) is achieved – equally, this effect is statistically significant in Model 06, which includes all variables.

Conversely, Table 03 shows that the coefficients for district magnitude are all negative and not significant statistically. Even so, these results are consistent, and they support the argument that electoral systems create incentives and affect party-group linkages (hypothesis I).

Our results for three organizational variables in all models (see Table 03 ) are in line with previous analyses of electoral systems, which supports the argument that party strength, intraparty cohesion, and financial linkage are positively associated with party-group linkages. In this sense, party organizational dimensions have a positive, statistically significant, and stable effect on party-group linkages over time.

In Model 02, where we examine party factors without our electoral system-level variables or control effects, we found variables that significantly affect party-group linkages. Parties with robust organizational structures, cohesive behavior from party elites, and strong financial linkages are more connected to social groups than those whose organization is fluid and weak. The relationship remains the same when controlling for the electoral system variables (Model 03), which indicates that aspects of party organization remain significant regardless of intraparty competition and contextual personalization incentives.

In models 02 and 03, those parties that make an effort to build solid extra-parliamentary party organizations forge more social ties than those focusing on alternative strategies over time, a finding that is consistent with our expectations and with previous studies on the topic ( COLEMAN, 1996 ; TAVITS, 2013 ). Models 02 and 03 also reveal – consistent with our expectations – a positive relationship between intraparty cohesion and linkages with non-party groups. By behaving cohesively over time, party members generate more stable and clear expectations from external players, including social groups. Because of stability and predictability, parties may have fewer incentives to break agreements. Additionally, they perform more effectively the fundamental functions of representation and accountability ( DALTON et al., 2011 ; ALDRICH, 2011 ).

In consonance with established literature, we found that financial linkage is positively associated with party-group linkages: a symbiotic relationship is formed. On the one hand, parties and candidates have a growing interest in increasing campaign funds. Conversely, non-party groups need party support to promote their policy agendas or to have access to electoral lists. This financial connection is probably a key factor in creating bonds between parties and groups ( BAWN et al., 2012 ; COX 1997 ). Nevertheless, contrary to our expectations, we did not find evidence suggesting that the degree of inclusiveness of candidate selection methods has significantly affected party-group linkages. Therefore, our results suggest that the rules for selecting candidates are not directly relevant, but we do not rule out their indirect effects (see Table 02 ).

Model 04 presents the effect of electoral systems and party organizations on party-group linkages, including all control variables described above: presidentialism index, division of power index, compulsory voting, civil society index, and educational equality. In essence, this is a re-estimation of Model 03 with added control variables. Once again, as the results in Model 04 show, the levels of party strength, financial linkages, and intraparty cohesion are, indeed, associated with increases in party-group linkages. The predicted effects are plotted in Figure 05 . Although the R-square of the model is relatively modest, all four graphs indicate that organizational and electoral system variables significantly affect party-group linkages – thus, our data seem to support hypothesis 02. In addition, and contrary to expectations, compulsory voting is the single control variable to have a significant and positive impact on party-group linkages. An alternative explanation is that increased participation possibly leads to increased electoral uncertainty, consequently stimulating the strengthening of group ties.

Figure 05
Predictive margins with 95% Cis (Model 04)

To what extent does a specific variation in the electoral system affect the relationship between parties and social groups? Do party structures in different electoral systems promote similar or different patterns of relations between parties and social groups? To systematically analyze how party agency plays out in different constraining institutional structures (electoral rules), we employ three models with the same variables used in Model 04 (excluding 'modified intraparty efficiency index'), but now we disaggregate our sample into different electoral systems: closed-list, flexible-list, and open-list/SNTV/STV ( MAINWARING, 1999 , p. 339)²⁶ .

Likewise, many have suggested that under preferential voting systems (flexible- and open-list systems), it is intuitively expected that party-level aspects would be irrelevant to explaining party-group linkages. However, as Table 04 shows, even the effects in Models 05 and 06 have declined vis-à-vis Model 07 (respectively, coefficients = 0. 501, 0.462, and 0.252, sig <0.05); party strength continues to have a significant impact on party-group linkages (see Figure 06 ).

Table 04
Results of panel analyses by party lists

Variables	(5)	(6)	(7)
	Closed-list system	Flexible-list system	Open-list/STV/SNTV systems
Party strength	0.501***	0.462***	0.252***
	(0.050)	(0.056)	(0.045)
Financial linkage	0.013	-0.08	0.010
	(0.040)	(0.058)	(0.044)
Inclusiveness in candidate selection processes	0.018	0.026	-0.081
	(0.037)	(0.05)	(0.059)
Internal cohesion	0.123***	0.036*	-0.007
	(0.019)	(0.019)	(0.024)
Magnitude (M)	-0.021	-0.004	-0.019
	(0.037)	(0.007)	(0.030)
Control variables	YES	YES	YES
Observations	726	444	698
Within R-squared	0.235	0.245	0.120
Number of parties	271	123	238
Parties FE	YES	YES	YES
Countries FE	YES	YES	YES
Regions FE	YES	YES	YES

• Source: Created by the author based on V-Party (2020) and V-Dem (2020) datasets .
• Notes: Standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1; dependent variable: party-group linkage index.

Figure 06
Predictive margins with 95% Cis (Model 07)

In contrast, Models 05, 06, and 07 confirm a solid linear pattern of decline in intraparty cohesion effects between different electoral systems: Closed-list and flexible-list systems are likely to favor the relationship between cohesive intraparty behavior and party-group linkages (with advantages to closed-lists – Models 05 and 06). However, the coefficient is not statistically significant if open-list systems are considered (STV and SNTV systems – Model 07). Therefore, our results indicate that each electoral system is likely (or not) to encourage different degrees of intraparty cohesion (CAREY and SHUGART, 1995; SHUGART, 2005, 2001). Moreover, the variables inclusiveness in candidate selection processes and financial linkage are not significant (see Table 04 ).

Notably, with respect to our findings, Brazil and Denmark are illustrative cases – both democracies allow 'preferential voting', adopting respectively OLPR and flexible-list PR electoral systems ( PASSARELLI, 2020 )²⁷ . In the Brazilian case, the increase in party organizational strength ( x̄ = -1.08 in 1995 to x̄ = -0.95 in 2018) is positively associated with a significant increase in party-group linkages (x̄ = -0,65 in 1995 to x̄ = -0,48 in 2018/r= 708, sig<0.00). Indeed, several recent studies have shown the relative strengthening of Brazilian party organizations (e.g., MAINWARING et al., 2017 ; RIBEIRO and LOCATELLI, 2019) . Similarly, and despite differences between the two cases, in Denmark party organizational strength (x̄ = -0.77 in 1990 to x̄ = -0.48 in 2018) is positively associated with party-group linkage densification (x̄ = -0.238 in 1990 to x̄ = -0.007 in 2018/r=0.842, sig<0.00) – in line with Allern et al. (2007) and Rahat and Kenig (2018) comparative evidence.

Overall, the connection between electoral systems and intraparty politics is not as automatic as usually expected. Our results suggest that party decisions about organizational strategies are also a significant factor. In line with many scholars who question the direct (or mechanical) relationship between social structure and party organizations (e.g., PRZEWORSKI and SPRAGUE, 1986 ; SARTORI 2005 ), our findings suggest that electoral systems have no determinant impact on the party-group relationships.

Combined with other contextual trends, maintaining or increasing linkages with society requires an organizational effort from parties that may or may not be directly stimulated by electoral rules. Therefore, our analysis seems to support Borz and Janda’s (2018) crucial argument: "political scientists are already moving away from a purely deterministic approach, whereby an organization is seen only as an effect of environmental conditions. The field of party politics should also consider parties' strategic decisions of adopting organizational changes to pursue their goals" ( BORZ and JANDA, 2018 , p. 01).

Conclusion

This article has offered a broad perspective on how and why institutions – particularly proportional (PR) electoral systems – are relevant to explain the relationship between parties as a collective enterprise and non-party groups. While much of the literature on political parties has focused on the socio-political determinants of parties’ decline or adaptation (e.g., BARTOLINI and MAIR, 2001 ; MAIR, 2013 ), we have shown how electoral institutions can negatively affect parties’ linkage strategies toward social groups over time, primarily by stimulating intraparty competition and personalization.

Nonetheless, we have also argued that the effects of electoral systems are linked to how party organizations are structured. Hence, party organizations are a critical intervening dimension in how the institutional environment affects party-group relationships – powerful, organized party elites are agenda-setters who have the final word in the organizational process ( BARNEA and RAHAT, 2007 ; COTTER et al., 1989 ; HARMEL and JANDA, 1994 ; TAVITS, 2013 ). V-Dem and V-Party data have allowed us to empirically investigate the longitudinal effects of party-level variables, which had not been systematically examined ( DÜPONT et al., 2022 , 2021 ).

In descriptive terms, we found that whereas average party-group linkages, party strength, inclusiveness in candidate selection processes, and intraparty cohesion remained stable over time (1989-2019), candidate-centred electoral systems have risen vis-à-vis a decline in financial linkages with social groups in the same period ( RAHAT and KENIG, 2018 ; RENWICK and PILET, 2016 ). Our models and empirical findings appear to demonstrate that strong and cohesive parties which receive financial support from society have consistently developed more linkages with non-party groups than those that are fluid and porous.

We also found evidence that political parties with strong organizations have established solid party-group linkages during this period, even in a candidate-centred electoral system (hostile environment). Indeed, this is consistent with our argument that party elites in strong party structures have more coordination capacity to mitigate intraparty conflicts. All things considered, our findings suggest that, even in more candidate-centred electoral systems, parties’ organizational structures may have more explanatory power than acknowledged in the extant literature: Parties’ organizational structures are highly capable of influencing party-group linkages.

In sum, our results suggest that adopting a more cautious analytical position on the relative balance between structure and agency is advantageous (or necessary), especially when it comes to party-society relationships ( GAUJA 2017 , pp. 13-15). Particularly, party-level implications, especially the organizational capacity of elites, should be carefully examined. In line with previous comparative research, our results suggest that the claim that "party change does not just happen" also holds true for parties’ relations with social groups ( HARMEL and JANDA 1994 , pp. 261-262). Finally, future analyses could examine if and to what extent institutional and party-level dimensions encourage clientelist or programmatic party-group linkages ( KITSCHELT, 2000 ; STROM, 1990) .

Acknowledgment

* We thank Juan Rodríguez-Teruel, Patricia Correa, and George Avelino for their suggestions on earlier drafts. A previous version of this paper was presented at the ECPR General Conference (2021). The usual disclaimer applies.

References

• ALDRICH, John H. (2011), Why parties? A second look. Chicago: University of Chicago Press. 400 pp..
• ALESINA, Alberto and SPEAR, Stephen E. (1988), An overlapping generations model of electoral competition. Journal of Public Economics. Vol. 37, Nº 03, pp. 359-379.
• ALLERN, Elin Haugsgjerd; OTJES, Simon; POGUNTKE, Thomas; HANSEN, Vibeke Wøien; SAURUGER, Sabine, and MARSHALL, David (2020), Conceptualising and measuring party-interest group relationships. Party Politics . Vol. 27, Nº 06, pp. 01-36.
• ALLERN, Elin Haugsgjerd; AYLOTT, Nicholas, and CHRISTIANSEN, Flemming Juul (2007), Social democrats and trade unions in Scandinavia: the decline and persistence of institutional relationships. European Journal of Political Research . Vol. 46, Nº 05, pp. 607–635.
• ALLERN, Elin Haugsgjerd and BALE, Tim (2012), Political parties and interest groups: disentangling complex relationships. Party Politics . Vol. 18, Nº 01, pp. 07–25.
• ALLERN, Elin Haugsgjerd and VERGE, Tània (2017), Still connecting with society? Political parties formal links with social groups in the twenty-first century. In: Organizing political parties: representation, participation and power. Edited by SCARROW, Suzan; WEBB, Paul D., and POGUNTKE, Thomas. Oxford: Oxford University Press. pp. 106–135.
• AMES, Barry (2001), The deadlock of democracy in Brazil . Ann Arbor: University of Michigan Press. 352 pp..
• ANDRÉ, Audrey; DEPAUW, Sam, and SHURGART, Matthew Søberg (2014) The effect of electoral institutions on legislative behaviour. In: Oxford handbook of legislative studies. Edited by MARTIN, Shane; SAALFELD, Thomas, and STRØM, Kaare W..Oxford: Oxford University Press. pp. 231–249.
• BALMAS, Meital; RAHAT, Gideon; SHEAFER, Tamir, and SHENHAV, Shaul (2014), Two routes to personalised politics: centralised and decentralised personalisation. Party Politics . Vol. 20, Nº 01, pp. 37–51.
• BARNEA, Shlomit and RAHAT, Gideon (2007), Reforming candidate selection methods: a three-level approach. Party Politics . Vol. 13, Nº 03, pp. 375–394.
• BARTOLINI, Stefano (2000), The political mobilization of the European Left, 1860–1980: the class cleavage. Cambridge: Cambridge University Press. 637 pp..
• BARTOLINI, Stefano and MAIR, Peter (2001), Challenges to contemporary political parties.In: Political parties and democracy . Edited by DIAMOND, Larry and GUNTHER, Richard. Baltimore: Johns Hopkins University Press. pp. 327-343.
• BAWN, Kathleen; COHEN, Martin; KAROL, David; MASKET, Seth; NOEL, Hans, and ZALLER, John (2012), A theory of political parties: groups, policy demands and nominations in American politics. Perspectives on Politics . Vol. 10, Nº 03, pp. 571–597.
• BENOIT, Kenneth and LAVER, Michael (2006), Party policy in modern democracies. LONDON: Routledge. 501 pp..
• BIZZARRO, Fernando; GERRING, John; KNUTSEN, Carl Henrik; HICKEN, Allen; BERNHARD, Michael; SKAANING, Svend-Erik; COPPEDGE, Michael, and LINDBERG, Staffan I. (2018), Party strength and economic growth. World Politics . Vol. 70, Nº 02, pp. 275–320.
• BORMANN, Nils-Christian and GOLDER, Matt (2013), Democratic electoral systems around the world, 1946–2011. Electoral Studies . Vol. 32, pp. 360–369.
• BORZ, Gabriela and JANDA, Kenneth (2018), Contemporary trends in party organization: revisiting intra-party democracy. Party Politics . Vol. 26, Nº 01, pp. 03–08.
• CAREY, John and SHURGART, Matthew Søberg (1995), Incentives to cultivate a personal vote: a rank ordering of electoral formulas. Electoral Studies . Vol. 14, Nº 04, pp. 417–439.
• CHANG, Eric C. C. and GOLDEN, Miriam A. (2007), Electoral systems, district magnitude and corruption. British Journal of Political Science . Vol. 37, Nº 01, pp. 115–137.
• CHHIBBER Pradee; JENSENIUS, Francesca Refsum, and SURYANARAYAN, Pavithra (2014), Party organization and party proliferation in Índia. Party Politics . Vol. 20, Nº 04, pp. 489–505.
• COLEMAN, John J. (1996), Party organizational strength and public support for parties. American Journal of Political Science . Vol. 40, Nº 03, pp. 805–824.
• COLOMER, Josep M. (ed) (2011), Personal representation: the neglected dimension of electoral systems. Colchester: ECPR Press. 216 pp..
• COPPEDGE, Michael; GERRING, John; KNUTSEN, Carl Henrik; LINDBERG, Staffan I.; TEOREL, Jan; MARQUARDT, Kyle L.; MEDZIHORSKY, Juraj; PEMSTEIN, Daniel; ALIZADA, Nazifa; GASTALDI, Lisa; HINDLE, Garry; PERNES, Josefine; von RÖMMER, Johannes; TZELGOV, Eitan; WANG, Yi-ting, and WILSON, Steven (2020), V-Dem varieties of democracy: methodology. Gothenburg: University of Gothenburg/ V-Dem Institute. 30 pp..
• COTTER, Cornelius P.; GIBSOB, James L.; BIBBY, John F., and HUCKSHORN, Robert J. (1989), Party organizations in American politics . Pittsburgh: University of Pittsburgh Press. 232 pp..
• COX, Gary W. (1997), Making votes count: strategic coordination in the world’s electoral systems. Cambridge: Cambridge University Press. 360 pp..
• DALTON, Russell J.; FARRELL, David M., and McALLISTER, Ian (2011), Political parties and democratic linkage: how parties organize democracy. Oxford: Oxford University Press. 272 pp..
• DALTON, Russell J. and WATTENBERG, Martin P. (eds) (2000), Parties without partisans: political change in advanced industrial democracies. Oxford: Oxford University Press. 328 pp..
• DÜPONT, Nils; KAVASOGLU, Yaman Berker; LÜHRMANN, Anna, and REUTER, Ora John (2022), A global perspective on party organizations. Validating the varieties of party identity and organization dataset (V-Party). Electoral Studies. Vol. 75. Available at ˂ https://doi.org/10.1016/j.electstud.2021.102423 ˃. Accessed on march, 25, 2021.
  » https://doi.org/10.1016/j.electstud.2021.102423
• DÜPONT, Nils; KAVASOGLU, Yaman Berker; LÜHRMANN, Anna, and REUTER, Ora John (2021), Party organizations around the globe: introducing the varieties of party identity and organization dataset (V-party). Working Paper . University of Gothenburg/ Varieties of Democracy Institute (V-Dem).
• FARRELL, David M. and McALLISTER, Ian (2006), Voter satisfaction and electoral systems: does preferential voting in candidate-centered systems make a difference? European Journal of Political Research . Vol. 45, Nº 05, pp. 723-749.
• FARRELL, David M. and SCULLY, Roger (2007), Representing Europe's citizens? Electoral institutions and the failure of parliamentary representation. Oxford: Oxford University Press. 200 pp..
• GAUJA, Anika (2017), Party reform: the causes, challenges and consequences of organizational change. Oxford: Oxford University Press. 214 pp..
• HARMEL, Robert K. and JANDA, Kenneth (1994), An integrated theory of party goals and party change. Journal of Theoretical Politics . Vol. 06, Nº 03, pp. 259–287.
• HAZAN, Reuven Y. and RAHAT, Gideon (2010), Democracy within parties: candidate selection methods and their political consequences. Oxford: Oxford University Press. 264 pp..
• IGNAZI, Piero (2012), Forza senza legittimità: Il vicolo cieco dei partiti. Roma: Laterza. 138 pp..
• JANDA, Kenneth (1980), Political parties: a cross-national survey. New York: The Free Press. 1019 pp..
• KATZ, Richard S. and MAIR, Peter (1995), Changing models of party organization and party democracy: the emergence of the cartel party. Party Politics . Vol. 01, Nº 01, pp. 05–28.
• KATZ, Richard S. and MAIR, Peter (eds) (1992), Party organizations: a data handbook on party organizations in western democracies, 1960-90. London: Sage Publications. 992 pp..
• KEY JR., Valdimer Orlando (1964) Politics, parties and pressure groups . New York: Thomas Y. Crowell Company. 738 pp..
• KITSCHELT, Herbert P. (2000), Linkages between citizens and politicians in democratic politics. Comparative Political Studies . Vol. 33, Nº 06, pp. 845–879.
• LAWSON, Kay and MERKL, Peter H. (eds) (1988), When parties fail: emerging alternative organizatons. Princeton: Princeton University Press. 606 pp..
• LÜHRMANN, Anna; DÜPONT, Nils; HIGASHIJIMA, Masaaki; KAVASOGLU, Yaman Berker; MARQUARDT, Kyle L.; BERNHARD, Michael; DÖRING, Holger; HICKEN, Allen; LAEBENS, Melis; LINDBERG, Staffan I.; MEDZIHORSKY, Juraj; NEUNDORF, Anja; REUTER, Ora John; RUTH-LOVELL, Saskia; WEGHORST, Keith R.; WIESEHOMEIER, Nina; WRIGHT, Joseph; ALIZADA, Nazifa; BEDERKE, Paul; GASTALDI, Lisa; GRAHN, Sandra; HINDLE, Garry; ILCHENKO, Nina; von RÖMER, Johannes; WILSON, Steven; PEMSTEIN, Daniel, and Brigitte SEIM (2020), Codebook varieties of party identity and organization (V–party) V1: varieties of democracy. (V–Dem) Project.
• MAINWARING, Scott P. (1999), Rethinking party systems in the third wave of democratization: the case of Brazil. Stanford: Stanford University Press. 412 pp..
• MAINWARING, Scott P.; POWER, Timothy J., and BIZZARRO, Fernando (2017), The uneven institutionalization of a party system: Brazil. In: Party systems in Latin America: institutionalisation, decay, and collapse. Edited by MAINWARING, Scott P.. New York: Cambridge University Press. pp. 164-200.
• MAIR, Peter (2013), Ruling the void: the hollowing of western democracy. London: Verso. 160 pp..
• MAYHEW, David R. (1974), Congress: the electoral connection. New Haven: Yale University Press. 216 pp..
• OLIVER, Christine (1991), Strategic responses to institutional processes. The Academy of Management Review . Vol. 16, Nº 01, pp. 145-179.
• PASSARELLI, Gianluca (2020), Preference voting systems: influence on intra-party competition and voting behaviour. London: Palgrave Macmillan. 286 pp..
• PEMSTEIN, Daniel; MARQUARDT, Kyle L.; TZELGOV, Eitan; WANG, Yi-ting; KRUSELL, Joshua, and MIRI, Farhad (2020), The V-Dem measurement model: latent variable analysis for cross-national and cross-temporal expert-coded data. V-Dem Working Paper Nº 21 . University of Gothenburg: Varieties of Democracy Institute.
• POGUNTKE, Thomas (2006), Political parties and other organizations. The handbok of party politics. Edited by KATZ, Richard S. and CROTTY, William. London: Sage. pp. 396–405.
• POGUNTKE, Thomas (2002), Party organization linkage: parties without firm social roots? In: Political parties in the new Europe: political and analytical challenges. Edited by LUTHER, Kurt Richard and MÜLLER-ROMMEL, Ferdinand. Oxford: Oxford University Press. pp. 43–62.
• POGUNTKE, Thomas; SCARROW, Suzan E.; WEBB, Paul D.; ALLERN, Elin H.; AYLOTT, Nicholas; van BIEZEN, Ingrid; CALOSSI, Enrico; LOBO, Marina Costa; CROSS, William P.; DESCHOUWER, Kris; ENYEDI, Zsolt; FABRE, Elodie; FARRELL, David M.; GAUJA, Anika; PIZZIMENTI, Eugenio; KOPEKÍ, Petr; KOOLE, Ruud; VERGE, Tània; MÜLLER, Wolfgang C.; SZCZERBIAK, Aleks; KOSIARA-PEDERSEN, Karina; RAHAT, Gideon, and van HAUTE, Emilie (2016), Party rules, party resources and the politics of parliamentary democracies: how parties organize in the 21st century. Party Politics . Vol. 22, Nº 06, pp. 661-678.
• PRZEWORSKI, Adam and SPRAGUE, John (1986), Paper stones: a history of electoral socialism. Chicago: University of Chicago Press. 224 pp..
• RAHAT, Gideon and KENIG, Ofer (2018), From party politics to personalized politics ? Party change and political personalization in democracies. Oxford: Oxford University Press. 400 pp..
• RENWICK, Alan and PILET, Jean-Benoit (2016), Faces on the ballot: the personalization of electoral systems in Europe. Oxford: Oxford University Press. 304 pp..
• RIBEIRO, Pedro Floriano and LOCATELLI, Luis (2019), Time after time: party organizational strength in new and old democracies. Opinião Pública . Vol. 25, Nº 01, pp. 199-233.
• SAMUELS, David J. and SHUGART, Matthew S. (2010), Presidents, parties, prime ministers: how the separation of powers affects party organization and behavior. Cambridge: Cambridge University Press. 310 pp..
• SARTORI, Giovanni (2005), Party types, organization and functions. West European Politics . Vol. 28, Nº 01, pp. 05-32.
• SCHMITTER, P (2001), Parties are not what they once were. In: Political parties and democracy . Edited by DIAMOND, Larry and GUNTHER, Richard. Baltimore: Johns Hopkins University Press. pp 67–89.
• SHURGART, Matthew Søberg (2005), Comparative electoral systems research: the maturation of a field and new challenges ahead. In: The politics of electoral systems. Edited by GALLAGHER, Michael and MITCHELL, Paul. Oxford: Oxford University Press. Pp. 25-26.
• SHUGART, Matthew Søberg (2001), Electoral ‘efficiency’ and the move to mixed-member systems. Electoral Studies . Vol. 20, Nº 02, pp. 173-193.
• STROM, Kaare (1990), A behavioral theory of competitive political parties. American Journal of Political Science . Vol. 34, Nº 02, pp. 565–598.
• TAVITS, Margit (2013), Post-communist democracies and party organization . Cambridge: Cambridge University Press. 295 pp..
• THOMAS, Clive S. (ed)(2001) Political parties and interest groups: shaping democratic governance. Boulder: Lynne Rienner. 353 pp..
• WATTENBERG, Martin P. (1994), The decline of American political parties, 1952–1992. Cambridge: Harvard University Press. 288 pp..
• WOOLDRIDGE, Jeffrey M. (2002), Introductory econometrics: a modern approach. Mason: Cengage Learning. 816 pp..

• *
  We thank Juan Rodríguez-Teruel, Patricia Correa, and George Avelino for their suggestions on earlier drafts. A previous version of this paper was presented at the ECPR General Conference (2021). The usual disclaimer applies.
• 2
  In organizational theory, Oliver (1991) proposed five strategies organizations use to respond to institutional conformation pressures. The author argued that these strategies reflect different levels of resistance to those pressures, ranging from passive to active strategy: acquiescence, compromise, avoidance, defiance, and manipulation.
• 3
  In general, we seek explanation for similarities in face of the many ‘things’ on which the cases differ ( JANDA, 1980 ).
• 4
  For an analysis at county level, see Bizzarro et al., (2018) .
• 5
  Other comparative efforts worth noting include Janda’s International Comparative Political Parties Project, Party Organizations Data Handbook ( KATZ and MAIR, 1992 ), and Political Parties Database ( POGUNTKE et al., 2016 ).
• 6
  Chang and Golden (2007 , p. 122) adopted the same criteria.
• 7
  Varieties of Party Identity and Organization (V-Party), version 1, October 2020. According to Lührmann et al. (2020 , p. 04), V-Party covers the 1970–2019 period and offers expert–coded assessments of party organization and identity for most relevant parties worldwide. In general, the V-party uses the same strategies as those in the V–Dem methodology: experts' evaluations. Altogether, 665 experts evaluated political parties' policy positions and organizational capacity across elections. As mentioned above, all parties selected have more than 05% of the vote share at a national election. The data were aggregated using V–Dem's Bayesian Item Response Theory measurement model ( PEMSTEIN et al., 2020 ).
• 8
  The end of the Cold War, notably marked by the Berlin Wall's fall in 1989, is considered the start point of the Fourth Wave of Democratization.
• 9
  For a relevant exception, see Thomas (2001, p , pp. 274-276). Instead of focusing on intraparty aspects, as we did, this author addressed the consequences of inter-party competition on party-group linkages.
• 10
  Affiliate organizations (v2pasoctie): “to what extent does this party maintain ties to prominent social organizations?” ( LÜHRMANN et al., 2020, p , p. 32).
• 11
  Party support group (v2pagroup): “ To which particular group in society does the core membership and supporters of this party belong?” ( LÜHRMANN et al., 2020, p , p. 31).
• 12
  Except the variable: ‘no specific, clearly identifiable group’ -- v2pagroup_0 (more details in Chart A01 - Appendix ). As a validation strategy, we found a positive Spearman correlation between v2pagroup (Affiliate Organizations/V-party 2020) and the sum of non-territorial sub-organizations (e.g., youth and woman branches) that are mentioned in the party statutes (official documents) (0.432, sig.< 0.05, n= 32), based on PPDB database ( POGUNTKE et al., 2016) . For more systematic robustness checks with different available datasets (e.g., JANDA, 1980 ; POGUNTKE et al., 2016) , see Düpont et al. (2022, 2021 , 2021 pp. 14-17).
• 13
  The literature is vast on this topic. Some of the most representative advocates of adaptation include Bartolini and Mair (2001) , Aldrich (2011) , and Dalton et al., (2011) ; among those who argue for the decline, see Dalton and Wattenberg (2000) , Schmitter (2001) and Mair (2013) . For a contemporary synthesis of this debate, see Rahat and Kenig (2018) .
• 14
  Similarly, Thomas (2001, p , p. 280) did not find a unique pattern in party-group linkages in cross-region comparative analyses, including Eastern Europe, Western Europe, and the Americas.
• 15
  According to Carey and Shugart (1995, p. 418): ‘in all systems where there is intraparty competition, as M grows, so does the value of personal reputation’.
• 16
  Data source: Bormann and Golder (2013) ; Carey and Shugart (1995); Chang and Golden (2007) ; Dalton et al., (2011) ; Farrell and Scully (2007) ; Passarelli (2020) ; Renwick and Pilet (2016) ; Shugart (2001) ; V-Dem (2020).
• 17
  For more details about the Colombian and Italian cases, see Passarelli (2020) .
• 18
  Party strength = local organisational strength ( v2paactcom ) + local party office (v2palocoff). About the data generation, it is important to note “all expert–coded variables s ave v2paactcom, v2paplur, v2parelig and v2pawelf converged according to strict V-Dem convergence criteria (no more than 10% of any set of parameters had Gelman-Rubin diagnostic values greater than 1.01). v2paactcom converged according to a weaker test of convergence (Gelman–Rubin diagnostic values greater than 1.1)” ( LÜHRMANN et al., 2020, p , p. 05).
• 19
  ‘Active community presence’ (v2paactcom): “To what degree are party activists and personnel permanently active in local communities?” ( LÜHRMANN et al., 2020, p , p. 32).
• 20
  ‘Local party offices’ (v2palocoff): “Does this party maintain permanent offices that operate outside of election campaigns at the local or municipal-level?” ( LÜHRMANN et al., 2020, p , p. 31).
• 21
  Financial linkage = large-scale donations from civil society (v2pafunds_3) + membership donations (v2pafunds_4).
• 22
  ‘Financial linkage’ (v2pafunds): “What were the major sources of party funds for this election campaign?”. Two answers are considered: “Large-scale donations from civil society organizations (including trade unions) (v2pafunds_3)/Membership fees and small-scale supporters' donations (v2pafunds_4)” ( LÜHRMANN et al., 2020, p , p. 34).
• 23
  Candidate nomination (v2panom): “Which of the following options best describes the process by which the party decides on candidates for the national legislative elections?” ( LÜHRMANN et al., 2020, p , p. 32-33).
• 24
  Elite cohesion (v2padisa): “To what extent do the elites in this party display disagreement over party strategies?” ( LÜHRMANN et al., 2020, p , p. 33).
• 25
  In short, $y_{it}=a+x_{it}b+v_i+e_{it}$ , where v_i(i=1, ..., n) are the fixed effects to be estimated. In addition, the Hausman Test was statistically significant (sig. <
• 26
  According to Mainwaring (1999): “with open list PR, with the single transferable vote, or with the single non-transferable vote, candidates owe their election to their own efforts, and are not beholden to the party” ( MAINWARING, 1999, p , p. 339).
• 27
  Denmark introduced the Flexible-list PR in 1920, while Brazil has employed OLPR since 1932.
• Data replication: https://doi.org/10.7910/DVN/A0YYKO

• Funding: Fundação de Amparo à Pesquisa do Estado de São Paulo Nº 2019/25865-0.

Annexes

Table A01 Description and statistics of variables

	N	Mean	Std. dev.	Min.	25 %	Median		75 %	Max.	V-dem variable codes
Party Strength	1,929	0.0	1.82	-4.03	-1.328	0.114		1.522	3.8	v2paactcom, v2palocoff
Party-group linkages	1,922	0.0	1.663	-3.412	-1.212	-0.124		1.166	6.008	v2pagroup, v2pasoctie
Financial linkages	1,920	0.0	1.504	-1.585	-1.85	-0.122		0.755	5.35	v2pafunds
Average District Magnitude (M)	1,930	29.5	56.8	2	5.29	9		15	250	v2elloeldm
Modified intraparty efficiency index	1,930	0.0	1.000	-1.073	-1.073	-0.015		1.043	2.101	v2elloelsy
Internal Cohesion	1,886	0.232	1.274	-3.072	-0.688	0.267		1.158	2.962	v2padisa
Inclusiveness in Candidate Selection Processes	1,930	0.503	1.419	-2.908	-0.405	0.580		1.439	4.659	v2panom
Presidentialism Index	1,930	0.155	0.174	0.009	0.036	0.083		0.23	0.931	v2xnp_pres
Division of power index	1,930	0.722	0.247	0.000	0.490	0.870		0.968	1.000	v2x_feduni
Compulsory Voting	1,930	0.530	0.831	0.000	0.000	0.000		1.000	3.000	v2elcomvot
Educational equality	1,930	1.50	1.50	-1.890	0.44	2.055		2.67	3.430	v2peedueq
Civil society index	1,930	0.872	0.121	0.000	0.830	0.900		0.950	0.980	v2xcs_ccsi

Source: V-Dem (2021) and V-Party (2020). Chart A01 Operationalization of variables

Variables
1) Party Strength = Local organisational strength + Local party office Local organisational strength (v2paactcom): To what degree are party activists and personnel permanently active in local communities? Clarification: please consider the degree to which party activists and personnel are active both during election and non-election periods. Party personnel refers to paid staff. Responses: 0: There is negligible permanent presence of party activists and personnel in local communities. 1: There is minor permanent presence of party activists and personnel in local communities. 2: There is noticeable permanent presence of party activists and personnel in local communities. 3: There is significant permanent presence of party activists and personnel in local communities. 4: There is widespread permanent presence of party activists and personnel in local communities . Local party office (v2palocof): Does this party maintain permanent offices that operate outside of election campaigns at the local or municipal-level? Clarification: By "local or municipal" we mean low level administrative divisions that are ranked below regions, provinces, or states. We refer to offices that ensure professional personal and continued interaction of the party with citizens. Permanent offices operate outside of election campaigns. Responses: 0: The party does not have permanent local offices. 1: The party has permanent local offices in few municipalities 2: The party has permanent local offices in some municipalities. 3: The party has permanent local offices in most municipalities. 4: The party has permanent local offices in all or almost all municipalities .
2) Financial linkage (v2pafunds) = Large-scale donations from civil society + Membership donations Party resources: What were the major sources of party funds for this election campaign? Clarification: choose up to three most important ones. If a main source of funding for this campaign comes from the party's assets such as properties and stocks, please code where these assets originally came from. Responses: 3: Large-scale donations from civil society organizations (including trade unions) [v2pafunds_3]. 4: Membership fees and small-scale supporters' donations [v2pafunds_4].
3) Internal Cohesion (v2padisa). Internal Cohesion: To what extent do the elites in this party display disagreement over party strategies? Clarification: Party strategies include election campaign strategy, policy stance, distribution of party financial resources, cooperation with other parties (i.e. coalition formation), and the selection of legislative and presidential candidates as well as the party leader. Party elites are prominent and influential party members such as current and former ministers, members of parliament or the party leadership, regional and municipal leaders, and opinion leaders. They do not necessarily have to be the part of the official party leadership. Responses: 0: Party elites display almost complete disagreement over party strategies and many party elites have left the party. 1: Party elites display a high level of visible disagreement over party strategies and some of them have left the party. 2: Party elites display some visible disagreement over party strategies, but none of them have left the party. 3: Party elites display negligible visible disagreement over party strategies. 4: Party elites display virtually no visible disagreement over party strategies.
4) Candidate Selection inclusiveness ( v2panom). Candidate nomination: Which of the following options best describes the process by which the party decides on candidates for the national legislative elections? Clarification: If nomination procedures vary across constituencies consider the most common practice. Responses: 0: The party leader unilaterally decides on which candidates will run for the party in national legislative elections. 1: The national party leadership (i.e. an executive committee) collectively decides which candidates will run for the party in national legislative elections. 2: Delegates of local/regional organisations decide which candidates will run for the party in national legislative elections. 3: All party members decide on which candidates will run for the party in national legislative elections in primaries/caucuses. 4: All registered voters decide on which candidates will run for the party in national legislative elections in primaries/caucuses
5) Party Society Linkages = Affiliate organisations + Party support group Affiliate organisations (v2pasoctie): To what extent does this party maintain ties to prominent social organisations? Clarification: When evaluating the strength of ties between the party and social organisations please consider the degree to which social organisations contribute to party operations by providing material and personnel resources, propagating the party's message to its members and beyond, as well as by directly participating in the party's electoral campaign and/or mobilisation efforts. Social organisations include: Religious organisations (e.g. churches, sects, charities), trade unions/syndical organisations or cooperatives, cultural and social associations (e.g. sports clubs, neighbourhood associations), political associations (e.g. environmental protection) and professional and business associations. Social organisations do not include paramilitary units or militias. Responses: 0: The party does not maintain ties to any prominent social organisation. 1: The party maintains weak ties to prominent social organisations. 2: The party maintains moderate ties to prominent social organisations. 3: The party maintains strong ties to prominent social organisations. Party support group (v2pagroup): To which particular group in society does the core membership and supporters of this party belong? Clarification: Choose only the key groups. Though you may choose up to three groups, if only one group is most relevant, please only choose that group. Responses: 0: No specific, clearly identifiable group [v2pagroup_0]. 1: The aristocracy, including high status hereditary social groups and castes [v2pagroup_1]. 2: Agrarian elites, including rich peasants and large landholders [v2pagroup_2]. 3: Business elites [v2pagroup_3]. 4: The military [v2pagroup_4]. 5: An ethnic or racial group(s) [v2pagroup_5]. 6: A religious group(s) [v2pagroup_6]. 7: Local elites, including customary chiefs [v2pagroup_7]. 8: Urban working classes, including labor unions [v2pagroup_8]. 9: Urban middle classes. (0=No, 1=Yes) [v2pagroup_9]. 10: Rural working classes (e.g. peasants) [v2pagroup_10]. 11: Rural middle classes (e.g., family farmers) [v2pagroup_11]. 12: Regional groups or separatists [v2pagroup_12]. 13: Women [v2pagroup_13]. 14: Other specific groups [v2pagroup_14].
6) Presidentialism Index (v2xnp_pres) Presidentialism Index: To what extent is the regime characterised by presidentialism? Clarification: Presidentialism means the "systemic concentration of political power in the hands of one individual who resists delegating all but the most trivial decision making tasks" (Bratton and Van de Walle, 1997: 63). It relates closely to V-Dem's index of Horizontal Accountability (v2x_horacc) but focuses more specifically on the extent to which the President is free from constraints by other institutions or actors. The point estimates for this index have been reversed such that the directionality is opposite to the input variables. That is, lower scores indicate a normatively better situation (e.g. more democratic) and higher scores a normatively worse situation (e.g. less democratic). Note that this directionality is opposite of that of other V-Dem indices, which generally run from normatively worse to better. Scale: Interval, from low to high (0-1).
7) Division of power index (v2x_feduni) Division of power index: Are there elected local and regional governments, and — if so — to what extent can they operate without interference from unelected bodies at the local level? Clarification: The lowest score would be reserved for a country that has no elected local or regional governments, or where all or nearly all elected offices are subordinate to non-elected offices at any local or regional level that exists. A high score would be accorded to a country in which both local and regional governments are elected and able to operate without restrictions from unelected actors at the local or regional level with the exception of judicial bodies. A medium score can be achieved in various ways: there are strong elected governments at the local level but not the regional level, or vice versa; or both local and regional governments elect an executive but not an assembly; or elected and non-elected offices are approximately equal in power at the local and regional levels; or various combinations of these scenarios. Scale: Interval, from low to high (0-1).
8) Compulsory voting (v2elcomvot) Compulsory Voting: Is voting compulsory (for those eligible to vote) in national elections? Responses: 0: No. 1: Yes. But there are no sanctions or sanctions are not enforced. 2: Yes. Sanctions exist and are enforced, but they impose minimal costs upon the offending voter. 3: Yes. Sanctions exist, they are enforced, and they impose considerable costs upon the offending voter.
9) Educational Equality (v2peedueq): Educational equality: To what extent is high quality basic education guaranteed to all, sufficient to enable them to exercise their basic rights as adult citizens? Clarification: Basic education refers to ages typically between 6 and 16 years of age but this varies slightly among countries. Responses: 0: Extreme. Provision of high quality basic education is extremely unequal and at least 75 percent (%) of children receive such low-quality education that undermines their ability to exercise their basic rights as adult citizens. 1: Unequal. Provision of high quality basic education is extremely unequal and at least 25 percent (%) of children receive such low-quality education that undermines their ability to exercise their basic rights as adult citizens. 2: Somewhat equal. Basic education is relatively equal in quality but ten to 25 percent (%) of children receive such low-quality education that undermines their ability to exercise their basic rights as adult citizens. 3: Relatively equal. Basic education is overall equal in quality but five to ten percent (%) of children receive such low-quality education that probably undermines their ability to exercise their basic rights as adult citizens. 4: Equal. Basic education is equal in quality and less than five percent (%) of children receive such low-quality education that probably undermines their ability to exercise their basic rights as adult citizens.
10) Core civil society index (v2xcs_ccsi) Core civil society index: How robust is civil society? Clarification: The sphere of civil society lies in the public space between the private sphere and the state. Here, citizens organise in groups to pursue their collective interests and ideals. We call these groups civil society organisations CSOs. CSOs include, but are by no means limited to, interest groups, labor unions, spiritual organisations if they are engaged in civic or political activities, social movements, professional associations, charities, and other non-governmental organisations. The core civil society index CCSI is designed to provide a measure of a robust civil society, understood as one that enjoys autonomy from the state and in which citizens freely and actively pursue their political and civic goals, however conceived. The index is formed by taking the point estimates from a Bayesian factor analysis model of the indicators for CSO entry and exit ( v2cseeorgs ), CSO repression ( v2csreprss ) and CSO participatory environment ( v2csprtcpt ). Scale: from low to high (0-1).
11) District Magnitude (M) (v2elloeldm) Lower chamber election district magnitude: For this election, what was the average district magnitude for seats in the lower (or unicameral) chamber of the legislature?
12) Electoral system index (v2elloelsy): Lower chamber electoral system — 13 categories: What was the electoral system used in this election for the lower or unicameral chamber of the legislature? Responses: 0: First-past-the-post (FPP, aka plurality) in single-member constituencies. The candidate with the most votes wins the seat. 1: Two-round system in single-member constituencies. Like FPP except that a threshold — usually 50% + 1 — is required to avoid a runoff between the two top vote-getters. 2: Alternative Vote in single-member districts. Voters rank-order their preferences for the candidates who compete for a single seat. If any candidate receives an absolute majority of first preferences, s/he is elected. If not, then the least successful candidates (based on first-preferences) are eliminated and their votes reallocated to the second-preferences. This process is repeated until a candidate reaches 50% +1 of the votes. 3: Block vote in multi-member districts. Electors have as many votes as there are seats within that district and can rank-order them (within or across parties) as they please. 4: Party block vote in multi-member districts. Voters cast a vote for a single party (but not for individual candidates within the party's list). The party with the most votes (i.e., a plurality) wins all the seats in that district. 5: Parallel (SMD/PR). Some seats are in single-member districts (allocated by FPP or two-round electoral rules) and other seats are in multimember districts (allocated by some form of PR). These districts are overlapping, meaning that each elector votes twice: once in the single-member district race and once in the multi-member district race. Results are independent. 6: Mixed-member proportional (SMD with PR compensatory seats). Some seats are in single-member districts (allocated by FPP or two-round electoral rules) and other seats are in multimember districts (allocated by some form of PR). These districts are overlapping, meaning that each elector votes twice: once in the single-member district race and once in the multi-member district race. Results are not independent. Specifically, the multimember seats are used to rectify disproportionalities achieved in the single-member district election — by adding seats, as necessary. This means that the representation of parties in the legislature is determined entirely by the PR ballot. It also means that the result of an MMP election is similar to the result of a PR election: parties achieve representation according to their nationwide vote share (on the PR ballot). 7: List PR with small multi-member districts (mean district size < 7). Each party presents a list of candidates for election within a district. Electors vote for a party, and parties receive seats in (rough) proportion to their overall share of the vote. Mean district size is less than seven. 8: List PR with large multi-member districts (mean district size > 7). Each party presents a list of candidates for election within a district. Electors vote for a party, and parties receive seats in (rough) proportion to their overall share of the vote. Mean district size is greater than seven. 9: Single-transferable Vote (STV) in multi-member districts. Electors rank-order candidates nominated for a district. Candidates that surpass a specified quota of first-preference votes are elected. The remaining seats are chosen by reallocating the votes of the least successful candidates to elector's second- (or third-) preferences until the specified quota is reached. This process is repeated until all seats for that district are filled. 10: Single non-transferable Vote (SNTV) in multi-member districts. Each elector chooses a single candidate. The candidates with the most votes (a plurality) win. (The number of winners is of course determined by the size of the district.) 11: Limited Vote in multi-member districts. Electors have more than one vote but fewer votes than the number of seats in the district. The candidates with the most votes (a plurality) win. (The number of winners is of course

Source: V-Dem (2020) and V-Party (2020).

Figure A01
Dependent Variable: Party-group linkage index

Notes: n=1930.

Table A02 Variables’ average by countries

	Countries	Party-group linkage	Party strength	Financial linkage	Average Magnitude (M)	Intraparty efficiency index	Candidate selection Inclusiveness	Intraparty cohesion	Number of Parties
1	Albania	-2.6	0.2	-1.6	11.7	-1.1	-2.1	-0.7	5
2	Argentina	0.8	0.5	0.0	5.4	-1.1	1.7	-0.4	18
3	Austria	0.5	0.7	0.3	5.5	0.0	0.9	0.7	8
4	Belgium	0.7	1.7	-1.6	12.7	0.0	2.7	1.1	16
5	Bolivia	0.2	-1.3	-0.4	14.4	-1.1	-0.8	0.3	8
6	Brazil	-0.6	-0.9	-0.4	19.0	1.0	0.4	0.3	18
7	Bulgaria	-1.1	0.3	-0.2	7.7	-1.1	-0.9	-0.2	16
8	Chile	0.2	-0.4	0.4	3.5	0.3	0.3	-0.1	10
9	Colombia	1.2	-0.8	0.0	4.9	1.5	-0.2	-0.7	11
10	Costa Rica	0.8	-2.8	-0.5	8.1	-1.1	0.3	-1.1	9
11	Croatia	-0.9	0.9	-0.1	12.6	-0.5	0.1	0.6	16
12	Cyprus	0.8	1.8	-1.3	9.3	1.0	0.8	0.0	7
13	Czech Republic	-0.7	-0.6	-0.6	18.0	0.0	1.0	-0.2	20
14	Denmark	0.0	-0.4	1.6	13.8	1.0	1.7	0.9	10
15	Dominican Republic	1.7	1.0	-0.1	4.7	0.1	1.2	-0.2	5
16	Ecuador	-1.0	-1.5	0.3	3.3	0.0	-1.6	0.6	27
17	El Salvador	1.4	0.9	0.0	4.8	-0.5	-0.2	0.0	7
18	Estonia	-1.5	0.2	-1.2	8.4	1.0	0.6	0.0	20
19	Finland	0.7	-1.1	1.5	13.6	1.6	2.1	1.5	10
20	Greece	1.2	1.7	0.4	5.2	0.4	-0.5	-0.2	10
21	Guatemala	-0.6	-2.2	-0.8	5.6	-1.1	-1.3	0.9	28
22	Honduras	2.5	1.0	0.6	7.1	0.1	1.6	0.2	5
23	Iceland	0.5	0.3	-0.3	8.5	0.0	1.6	1.2	11
24	Ireland	-0.6	1.5	0.9	4.0	2.1	1.4	1.3	6
25	Israel	-0.2	-0.1	-1.6	120.0	-1.1	0.0	0.2	17
26	Italy	-0.6	-0.1	0.2	21.9	-0.6	1.4	-0.1	22
27	Japan	1.2	1.3	2.3	4.0	2.1	0.0	0.3	6
28	Kosovo	0.5	0.7	0.2	60.0	1.0	-0.8	0.1	10
29	Latvia	-0.7	0.2	0.2	20.0	0.0	0.1	0.2	28
30	Luxembourg	-1.2	-1.1	0.2	15.0	1.0	1.3	1.7	6
31	Malta	1.4	0.9	1.1	5.0	2.1	-0.4	-0.2	2
32	Montenegro	0.6	1.6	-1.2	81.0	-1.1	-0.6	0.7	16
33	Netherlands	-1.0	-3.0	1.1	150.0	0.0	1.3	0.8	8
34	Nicaragua	1.8	-0.1	-0.7	5.5	-1.1	-0.3	0.6	7
35	Norway	-1.5	0.9	-1.1	8.4	-1.1	1.3	-0.2	7
36	Paraguay	0.6	-0.1	-1.2	4.4	-1.1	2.5	-0.5	8
37	Peru	-1.1	-2.2	0.2	5.6	1.0	-1.4	0.9	25
38	Poland	0.1	0.3	1.0	10.8	1.4	0.1	-0.3	23
39	Portugal	1.2	1.7	0.2	10.5	-1.1	0.8	-0.9	5
40	Romania	-2.0	0.4	-0.7	7.7	-1.1	0.1	0.8	11
41	Serbia	-0.8	-0.3	-0.1	250.0	-1.1	-1.0	0.1	29
42	Slovakia	-1.2	-0.3	-1.0	150.0	0.0	-0.1	-0.8	23
43	Slovenia	0.7	1.0	0.2	10.0	0.8	1.1	0.3	16
44	Spain	0.5	0.9	0.9	6.8	-1.1	0.7	0.3	11
45	Sweden	-0.6	0.1	-1.0	11.7	0.0	1.3	0.0	9
46	Switzerland	2.3	1.0	5.4	7.7	1.0	1.3	-	17
47	Ukraine	1.5	0.7	0.0	450.0	-1.1	-0.7	0.5	5
48	Uruguay	1.6	1.2	-1.1	5.2	-1.1	1.8	-1.0	5

Source: V-Dem (2021) and V-Party (2020).

Notes: n=1930.

Edited by

• Revised by Karin Blikstad

Publication Dates

• Publication in this collection
  19 Sept 2022
• Date of issue
  2022

History

• Received
  02 Aug 2021
• Accepted
  18 Apr 2022