Explaining legislative party switching in advanced and new democracies

Abstract

The phenomenon of legislative party switching has attracted considerable attention among scholars. Studies show that parties are likely to experience defections if they are unable to offer their legislators electoral, office and policy benefits. While often deemed a symptom of a weakly institutionalized party system, switching may also be frequent in established democracies. Yet, surprisingly, we have no empirical evidence to help us understand whether party benefits that drive switchers are different in established democracies and new democracies. I argue that parties that fail to provide their members with re-election prospects or government access are likely to witness switching only in new democracies. In advanced democracies, switching is likely to be associated with ideological identity and policy motives. The statistical analysis of party switching across 25 European advanced and post-communist democracies confirms my expectations.

Keywords

electoral volatility legislative studies party switching party system institutionalization programmatic structuring

Political parties are at the core of representative democracy as they serve as a mechanism for competition among elites for the people’s vote (Dahl, 1971; Downs, 1957; Schumpeter, [1942] 2013). By enabling elites to compete, political parties constitute their main access to power; they recruit, select and socialize politicians for election (Norris, 1997). Once elected, politicians use their parties strategically to obtain various benefits. When their parties cease to provide them with those benefits, they may defect and switch to a party that better furthers their goals (Aldrich, 1995; Aldrich and Bianco, 1992). Given the significance of parties in a legislator’s career, it is not surprising that the phenomenon of party switching has been studied extensively in recent years.

The main body of party switching research focuses on the factors that drive legislators to change their party affiliation. Building on ambition theory, scholars have demonstrated that members of parliament (MPs) may switch parties to secure re-election, obtain office benefits and gain policy pay-offs (e.g. Desposato and Scheiner, 2008; Heller and Mershon, 2005; McElroy, 2003; McMenamin and Gwiazda, 2011; Reed and Scheiner, 2003; Thames, 2007). Most research on switching looks at new or non-Western democracies, suggesting that party switching is more common in weakly institutionalized political systems where parties are vaguely defined and convey ambiguous information about their policy goals (Desposato, 2006; Kreuzer and Pettai, 2003; McMenamin and Gwiazda, 2011; Shabad and Slomczynski, 2004; Young, 2014). In such systems, politicians use parties as vehicles to secure gains without risking electoral and ideological costs. However, party switching does not only occur in weak party systems (Heller and Mershon, 2005, 2008; Kato, 1998; Kato and Yamamoto, 2009; Klein, 2018; Mershon and Shvetsova, 2013a; Nokken, 2009; Nokken and Poole, 2004; O’Brien and Shomer, 2013; Reed and Scheiner, 2003), but surprisingly we still do not know whether party switchers in established democracies are driven by the same incentives as their counterparts in weakly consolidated ones.

I argue that while politicians do seek to promote their electoral, office and policy benefits in both new and advanced democracies, the utility from obtaining those different goals varies. In institutionalized and established democracies, party systems are programmatically structured and tend to result in considerable transaction costs for switchers, thus dwarfing the impact of vote and office incentives on switching. When party labels are ideologically defined and distinguishable, switchers are very likely to find themselves members of an ideologically incompatible party. As a result, they will be unable to promote their own agenda and be reluctant to toe the party line. This can upset potential voters and incur both electoral and ideological costs (Desposato, 2006). However, parties in new democracies are not arranged along a deeply embedded set of ideological positions, and the utility of holding office is less likely to be a function of ideology. Thus, while switching tends to be mainly vote- and office-driven in new democracies, it is likely to be based on policy grounds in established democracies.

To test the argument, I look at party switching in Europe, distinguishing between advanced and new democracies. Drawing on a new data set across 25 countries, I test the three motivational predictors of switching: parties’ ability to offer their members vote, office and policy benefits. The results are consistent with the hypotheses and provide four main insights to our understanding of party switching and voter–representative linkages. First, the findings suggest that the link between party switching and vote and office incentives applies only to weak party systems. In such systems, unattractive parties, which are incapable of delivering seats and office benefits to their candidates, tend to witness switching. In established democracies, party switching is associated with policy motivations rather than electoral and office benefits offered by a candidate’s party. Second, drawing on survey data, I show that switching is affected by the extent to which a party system is programmatically structured. Third, the results reveal that party switching in consolidated democracies may be a product of a volatile electorate. This is not the case in new democracies, where fluctuation among the political elite is not triggered by an unstable electorate. Last, the findings join existing studies (Klein, 2018; Nielsen et al., 2018) and give rise to questions about the perceived misrepresentation of party switching. The article suggests that representatives who change their party affiliation might reflect voters’ will to a much greater extent than usually claimed.

Party switching: Costs and benefits

Scholarship on the determinants of party switching is heavily based on the notion that politicians are ambitious. One of the most fundamental models of legislative behaviour (Müller and Strøm, 1999) suggests that political agents seek to maximize (1) their political office benefits, (2) their influence on public policy and (3) their votes.¹ Politicians strategically use their parties to fulfil these ambitions. Indeed, it is argued that the formation of political parties aimed to overcome problems of collective action as they reduce the cost for individual politicians of the pursuit of votes, office and policy benefits. In addition, politicians use their party’s label to signal their policy position to voters, who are consequently more capable of making informed decisions. As a result, parties help individuals to associate legislative achievements with party labels, thus securing electoral rewards for individual MPs (Aldrich, 1995; Cox and McCubbins, 2005, 2007). Politicians may use their party’s electoral, office and policy resources as long as it maximizes their utility. When it does not, they may defect and join a party that provides them with greater benefits (Aldrich, 1995).

In fact, it has been argued that electoral incentives are one determinant of party switching: once legislators believe that their prospects for re-election are low, they may switch to a party that enhances their chances of re-election. As many studies have shown, politicians may leave losing parties, that is, parties that are likely to experience losses in vote share, when compared with their performance in the last election (Desposato and Scheiner, 2008; Desposato, 2006; Reed and Scheiner, 2003; Thames, 2007; Young, 2014), in the coming election (O’Brien and Shomer, 2013), and as reflected as well by those parties’ popularity in opinion polls (McMenamin and Gwiazda, 2011).

Parliamentarians may also use their party for office benefits. Studies have shown that incumbents seek to obtain an office that is more important than the one they are holding in their party, in parliament or in the executive branch. They may therefore defect to parties that can offer those benefits (Desposato, 2006; Heller and Mershon, 2005; Kato, 1998; Laver and Benoit, 2003; McLaughlin, 2011; Thames, 2007; Young, 2014). Moreover, Heller and Mershon (2008) have shown that strong party discipline creates incentives to switch because it may hamper MPs’ pursuit of one or more goals (higher office and even re-election).

Legislators also seek to attain policy goals and to improve their ideological pay-offs by switching parties. Politicians defect from parties that impede the implementation of their ideologically inspired policy goals, namely because the party’s ideal point is far from their own (e.g. Herron, 2002; McElroy, 2003; Thames, 2007). Likewise, O’Brien and Shomer (2013) have shown that legislators are less likely to switch from ideologically cohesive parties.

Legislators who contemplate changing party affiliation consider the possible costs of switching. As mentioned, like-minded candidates form political parties to advance their agenda and obtain ideological benefits. Legislators who defect from their parties to get re-elected may later find that their new party’s prevailing ideology is incompatible with their own and that the party is less capable of implementing their ideology.² In addition to the ideological cost, ideological divergence may translate into electoral costs for switchers since it may upset both their voters and party leaders (Desposato, 2006). Moreover, in order to obtain a seat in parliament, political candidates take advantage of their parties’ resources, such as party activists, well-funded campaigns and effective organizations. Legislators who switch parties often have to re-socialize in their new parties, thus relinquishing the electoral resources that got them into parliament in the first place.

Finally, voters prefer that their representatives are loyal to the party under whose label they have been elected. Voters greatly value clear party labels; they use them as information cues and seek to prevent any distortion of those labels (Mershon and Shvetsova, 2013a, 2013b; Sevi et al., 2018). Voters may therefore hold a grudge against candidates who switch parties, and the possibility of being electorally punished by voters has ‘the capacity to put a brake on the MP’s impulse to defect’ (Mershon and Shvetsova, 2013b: 37).

Programmatic party competition in new and established democracies and party switching

Having laid out the inducements and deterrents to party switching, I will now discuss the programmatic structuring of party competition and how it may affect the electoral and ideological deterrents to switching. Party competition is programmatically structured if ‘teams of politicians compete for votes by offering citizens alternative packages of policies that they commit to enact if elected to political office’ (Kitschelt et al., 2010: 16). According to this approach, voters assess the candidates’ perceived policy ideal points and choose a candidate who best promotes their programmatic interest. Hence, in programmatic party competition voters develop attachments to parties on the basis of ideological congruence.

In order to efficiently attract voters, candidates in ideologically structured party systems are organized under clear party labels that distinguish themselves from one another by their programmatic packages (Kitschelt, 2000). In such systems, clear party labels help to form partisans’ opinions (Zaller, 1992). Arguably, party cues reduce the cost of attaining information about a party’s policy objectives. When a party possesses distinctive ideological preferences, its supporters receive information cues and are able to better adopt its positions. Furthermore, clear ideological labels encourage stronger programmatic agent–principal linkages between voters and their representatives (Kitschelt et al., 2010; Mainwaring and Scully, 1995).

Party competition in established democracies is ideologically structured: parties are well rooted in society, and citizens have developed loyalty to parties (Cox, 1997; Mair, 1997). However, party systems in new democracies are still in the process of institutionalization. Party behaviour is inconsistent, the differences between parties’ ideologies are ambiguous, and parties do not possess clear and distinctive ideological labels (e.g. Kitschelt et al., 1999; Mainwaring and Scully, 1995; Mainwaring and Torcal, 2006; Przeworski, 1975; Rose, 1995; Tavits and Letki, 2009).

The extent to which a party system is programmatically structured is consequential for party switching. Established democracies, which exhibit high levels of programmatic party competition, tend to have higher defection-deterring costs. When parties distinguish themselves ideologically from one another, switchers are less able to promote their agenda in the receiving party that is ideologically incompatible. However, in cases where the party changes its policy position or does not implement the policy promised to voters during the election campaign, defections are less likely to be ideologically costly. In other words, when politicians consider switching on policy grounds, we can expect the receiving party³ to be more ideologically compatible with them than the one from which they have defected. However, in new democracies, the party system is not strongly programmatically structured, and party labels are not clearly defined. This enables switchers to secure vote and office benefits without suffering ideological setbacks.

Legislators in advanced democracies also face greater electoral costs. As party labels are much more valued by voters in such systems, voters prefer individual politicians to be loyal to their party labels. Thus, legislators face defection-deterring costs because voters rely on party labels and are keen to prevent the destruction or erosion of those labels (Mershon and Shvetsova, 2013a, 2013b). Yet switchers are less likely to turn off constituencies in weakly institutionalized democracies, where voters themselves show low levels of loyalty.

To summarize, I argue that while political parties provide legislators with seats, posts and policy benefits in both new and advanced democracies, the expected cost of party switching will alter the utility magnitude from obtaining those benefits. Advancing electoral and office benefits by switching is much less beneficial for politicians in programmatically constructed party systems, as the ideological and electoral costs will diminish MPs’ net utility. Yet politicians seeking to promote their ideological pay-offs by switching are likely to succeed in advanced democracies. This is not the case in new democracies, where party switching is unlikely to increase politicians’ policy utility.

H1: Parties that cannot deliver re-election prospects or office benefits to their members are likely to witness switching mainly in new democracies.

H2: Parties that cannot deliver policy pay-offs to their members are likely to witness switching mainly in advanced democracies.

To further investigate the mechanism of the effect, I examine how the extent to which a party system is programmatically structured affects switching. I expect that the defection-deterring costs of switching will grow stronger in systems with programmatically structured party competition. However, switchers will be less likely to face those costs in party systems where party competition is not ideologically crystallized.

H3: The more programmatically the party competition is structured, the less likely parties are to witness switching.

Data and research methodology

To test the hypotheses, I have collected data on legislative inter-party switches⁴ and political institutions from 25 Western and post-communist European democracies over the period 1990–2013. This case selection is justified because it provides sufficient variation on the explanatory variables along the party and country dimensions.

I use party-level observation⁵ as the unit of analysis. Although this study explores the phenomenon of individual legislative switching, party-level analysis is more suitable for this cross-national study, as the key covariates are at the party level and cannot explain individual-level variation.

The dependent variable accounts for the number of legislative defections a given party witnesses in a given parliamentary term. I record⁶ any MP⁷ who defected from the party under whose label she or he had been elected as a switch, which provides an absolute number of switches for a given party in a parliamentary term.

Measuring vote incentives

To capture the extent to which parties provide their MPs with re-election prospects, the vote share variable accounts for legislators’ assessment of their parties’ probability of gaining votes in the next election. I measure the difference between a party’s vote share in the upcoming election (t + 1) and the party’s vote share in the election that established the sitting parliament (O’Brien and Shomer, 2013). When a party’s vote share is likely to increase, its MPs are likely to have assessed their probability of being re-elected as higher than legislators of a losing party.⁸

Measuring office incentives

Public offices, as well as the most important parliamentary offices, are mainly controlled by the government and less so by the opposition. Thus, maximization of office pay-offs is more likely to be found in coalition parties. I score the variable government as a continuous variable, measuring the number of months a party was part of the coalition (as a percentage of the total months in a given parliamentary term).

Measuring policy incentives

To date, the testing of policy motivations for party switching has mainly stressed the ideological differences between the switcher and her or his party as measured by legislative voting in parliament (e.g. Desposato and Scheiner, 2008; Desposato, 2006; McElroy, 2003). Others have measured the degree to which the party is ideologically cohesive (O’Brien and Shomer, 2013) as a proxy for policy motivations. For a cross-national analysis (using aggregate measures), Rice scores for ideologically cohesive parties could be a good solution but may, as O’Brien and Shomer (2013) note, give biased estimates because the scores cannot differentiate between cohesion and discipline and do not account for absences and abstentions.

This study uses data provided by the Comparative Manifesto Project (CMP) to measure policy incentives for party switching. It does so by capturing shifts in party policy as recorded in the CMP (Volkens et al., 2014) for the one-dimensional left–right ideology scale of each party. Although the CMP has been confronted with critique,⁹ no other source measuring parties’ policy positions is as systematic and extensive in their temporal and spatial coverage as the CMP (Gemenis, 2013).

Since I am interested in the magnitude of the policy change and not its direction, the variable policy accounts for the number of points by which the party policy changed between the upcoming and the previous election. A considerable policy shift indicates that a party’s ideology has changed considerably during the parliamentary term. Policy-driven switching may thus occur when legislators are not satisfied with a major shift in policy.

Measuring programmatic structuration

The concept of programmatic structuring of party competition implies that individuals cast their votes on the basis of ideology for parties that possess clearly defined labels. I follow Mainwaring and Torcal’s measure (2006) to assess the degree of perceptual disagreement in a given country at the time of the last election. I employ the Comparative Study of Electoral System (CSES), the Central & Eastern Eurobarometer Trend File and the European Social Survey (ESS) and construct the score by using the distribution of the respondents’ placements of their favourite parties.¹⁰ I score a country by calculating the mean standard deviation of all parties, weighted by the number of party supporters. A high score indicates higher levels of mixed understanding on the part of voters in a given country about national parties’ left–right position on the ideological axis, thus suggesting low levels of programmatic structuring of the party system.

Controls

Electoral volatility: An extensive literature has examined the linkages between electoral volatility and party system instability. One argument holds that it is the behaviour of political elites that determines the stability of the electorate in a top-down process. According to this school of thought, instability on the supply side, that is, on the side of parties and their members, triggers voting fluctuation among the citizens. When the elite level is not consolidated, there is a lack of strong party organization and clear policy platforms, which makes voters reluctant to develop attachment to parties (McMenamin and Gwiazda, 2011; Powell and Tucker, 2014; Rose et al., 2001). As Mershon and Shvetsova (2013b) have demonstrated, inter-party movements by MPs can lead to electoral volatility. However, once political actors present more stable and consistent alternatives, voters become more likely to develop coherent voting patterns (Tavits, 2008). Another argument reverses the causal direction and focuses on the demand side as the driving force for the consolidation of party systems. The key point is that electoral dealignment leads to instability in party systems: party competition should be stable as long as voters maintain their loyalty to existing parties (e.g. Butler and Stokes, 1974; Mainwaring and Torcal, 2006; Taagepera and Grofman, 2003; Tavits, 2005). Moreover, it is worth noting that electoral volatility in new democracies does not implicate the same voter behaviour as in established democracies. In post-communist Europe, electoral volatility reflects inexperienced voters whose behaviour is a response to, rather than a cause of, an unstable party system (Tavits, 2008). This is not the case in institutionalized democracies where electoral volatility represents the behaviour of sophisticated voters who rely less on old partisan loyalties and cast their votes in an informed way (Van der Meer et al., 2015). This study employs the Pedersen Index (Pedersen, 1983) to capture the aggregate turnover from one party to others between the prior and current (t − 1) election.¹¹

Electoral systems: Studies also show that the extent to which switchers face electoral costs is dependent on the electoral rules, without consistent findings nonetheless. For example, in electoral systems where party labels are able to shield switchers from electoral punishment (i.e. party-centred systems), politicians are more inclined to switch (Heller and Mershon, 2005; Klein, 2016; Mershon and Shvetsova, 2013a). Moreover, electoral laws can affect the way party leaders strategically attract switchers, depending on their ability to nominate the candidates (Kato and Yamamoto, 2009). On the other hand, scholars showed how candidate-centred electoral systems may encourage switching, as the greater incentives to cultivate a personal vote (Carey and Shugart, 1995) lead politicians to rely more on local supporters and less on their party for their re-election (Klein, 2018; O’Brien and Shomer, 2013; Thames, 2016). I include the dummy variable open ballot (Carey and Shugart, 1995), which indicates whether the ballot structure allows voters to express preferences among candidates who run under the same party labels (1) or not (0). The variable district magnitude accounts for the average of the district magnitudes in the country for a given parliamentary term.¹² To further control for personal vote effects, I introduce an interactive term of open ballot × district magnitude, following previous studies on the electoral incentives for personal vote (André et al., 2012, 2015; Carey and Shugart, 1995). District magnitude is argued to have a differential effect. Greater district magnitude increases incentives for personal vote when ballot structure is open, and decreases the incentives when ballot structure is closed.

Effective number of parties: A high number of receiving parties might increase the likelihood of switching. I control for this by measuring the effective number of parties, that is, the number of parties weighted according to the proportion of seats.

Duration: The duration of the term may positively influence the number of switches, so the model controls for the duration (measured in months) of the parliamentary term (Mershon and Shvetsova, 2013a, 2013b).

Party size: Since party size may affect the prevalence of switching, I control for the party’s seat share in the house.

Because the dependent variable is a count of the switches that occurred during a parliamentary term, a Poisson model should be employed. Nevertheless, as Table 1 shows, the assumption of the Poisson model that the standard deviation of the count is equal to the mean is violated. Therefore, I use a negative binomial model with varying-intercept random effects. The model analyses 505 party observations that are nested in 87 country-parliamentary terms.

Table 1.

Descriptive statistics of the variables.

	Min	Mean	Max	SD
Switching	0	2.4	79	7.2
Vote share	0	1.08	6.08	0.65
Government	0	0.36	1	0.46
Policy	0	10.9	65.3	10.8
Perceptual disagreement	1.08	1.67	2.55	0.27
Open ballot	0	0.62	1	0.48
District magnitude	1	3.38	4	0.89
ENP	2	4.23	10.8	1.46
Duration	7	44.98	70	10.67
Electoral volatility	1.73	23.3	100	19.24
Party size (ln)	0	4.01	6	1.31

Note: Descriptive statistics of the variables. ENP = effective number of parties; SD = standard deviation.

Results and discussion

Table 2 presents the frequency of party switching among the parties included in the study. On average, 57.8% of parties in post-communist democracies have witnessed switching compared to only 30.75% of parties in advanced countries. In some countries, most of the observed parties have exhibited switching (67% of the parties in the Czech Republic and Romania, 60% in France, 75% in Italy and 92% in Poland), in others only few (about 10% of the parties in Belgium, Norway and the Netherlands). The variation is great also at the individual level; every seventh legislator switches in Romania, and every tenth in Italy, Latvia and Poland, compared to less than 1% in Belgium, Germany, Greece, Spain, Sweden and the United Kingdom.

Table 2.

Descriptive statistics of the dependent variable.

	Time period (terms)	Observations (parties)	Party switchers
	Time period (terms)	Observations (parties)	(Parties)	(MPs)
Austria	1995–2008 (4)	17	3 (18%)	26 (3.6%)
Belgium	1995–1999 (1)	10	1 (10%)	1 (0.7%)
Czech R.	1992–1998, 2002–2013 (5)	21	14 (67%)	89 (15.1%)
Denmark	2001–2011 (3)	23	5 (22%)	11 (2.1%)
Estonia	2003–2011 (2)	8	6 (75%)	14 (9%)
Finland	1991–2011 (5)	40	12 (30%)	18 (1.8%)
France	2002–2012 (2)	10	6 (60%)	20 (1.8%)
Germany	1990–2013 (6)	30	13 (43%)	17 (0.4%)
Greece	1990–2004 (4)	14	5 (36%)	11 (0.9%)
Hungary	1994–2010 (4)	18	10 (56%)	89 (6%)
Iceland	1999–2007 (2)	10	3 (30%)	4 (3.2%)
Ireland	1997–2012 (3)	18	5 (28%)	18 (3.8%)
Italy	1992–2013 (6)	44	33 (75%)	349 (11.5%)
Latvia	1993–1995, 1998–2002 (2)	8	4 (50%)	16 (10.8%)
Lithuania	1992–1996 (1)	5	1 (20%)	1 (0.8%)
Netherlands	1994–2010 (5)	38	4 (11%)	8 (1.1%)
Norway	1993–2009 (4)	28	3 (11%)	10 (1.5%)
Poland	1991–2011 (6)	24	22 (92%)	271 (13.2%)
Portugal	2005–2009 (1)	6	2 (33%)	2 (0.9%)
Romania	1996–2000, 2004–2012 (3)	12	8 (67%)	159 (14.5%)
Slovakia	2002–2012 (3)	16	8 (50%)	36 (8.1%)
Slovenia	2000–2011 (3)	21	9 (43%)	23 (8.8%)
Spain	1993–2011 (5)	44	6 (14%)	10 (0.6%)
Sweden	1998–2010 (3)	21	5 (24%)	6 (0.5%)
United Kingdom	1992–2010 (4)	19	9 (47%)	21 (0.8%)
		505	1230

Note: Descriptive statistics of the dependent variable. MPs = members of parliament.

Table 3 presents the negative binomial regression results for the effect of numerous predictors on the frequency of party switching in established democracies (models 1–2) and in post-communist democracies (models 3–4).¹³ As model 1 demonstrates, vote-seeking incentives cannot explain party switching in advanced democracies. The variable vote share is not statistically significant, which suggests that a party’s ability to provide its members with re-election prospects does not affect its probability of witnessing switching. These results indicate that the evidence of vote-driven switching found in previous research is not present in the case of advanced democracies. The model also reveals that the typical results concerning office-driven switching, which are based on existing scholarship, do not apply to advanced democracies. The variable government is not statistically significant, indicating that government parties do not have a greater expectation of losing members than opposition parties. Hence, parties’ ability to offer their MPs office benefits does not have an impact on switching.

Table 3.

Random effect negative binomial models of Party Switching.

	Advanced		Post-communist		All
	(1)	(2)	(3)	(4)	(5)	(6)
Vote share	−0.168 (0.199)	−0.157 (0.199)	−0.710 (0.200)***	−0.697 (0.205)***	−0.450 (0.147)**	−0.449 (0.147)**
Government	−0.232 (0.276)	−0.233 (0.277)	−0.732 (0.325)**	−0.749 (0.322)**	−0.354 (0.208)	−0.354 (0.208)
Policy	0.021 (0.010)*	0.022 (0.010)**	0.019 (0.014)	0.019 (0.014)	0.018 (0.008)*	0.018 (0.008)*
Electoral volatility (t − 1)	0.046 (0.016)**	0.046 (0.016)**	−0.008 (0.013)	−0.013 (0.013)	0.028 (0.008)***	0.029 (0.008)***
Perceptual disagreement	0.790 (0.713)	0.781 (0.704)	0.961 (0.762)	0.320 (0.743)	1.270 (0.466)**	1.273 (0.463)**
District magnitude	−0.024 (0.193)	0.067 (0.204)	−0.215 (0.318)	0.029 (0.306)	−0.034 (0.168)	0.035 (0.182)
Open ballot	−0.220 (0.348)	1.059 (1.160)	−0.171 (0.578)	4.825 (2.444)*	−0.229 (0.276)	0.780 (1.116)
ENP	0.092 (0.140)	0.102 (0.137)	0.195 (0.171)	0.255 (0.167)	0.049 (0.116)	0.049 (0.115)
Duration	0.026 (0.015)	0.024 (0.015)	−0.003 (0.021)	−0.013 (0.019)	0.012 (0.013)	0.010 (0.013)
Party size (ln)	0.543 (0.101)***	0.542 (0.101)***	0.793 (0.181)***	0.733 (0.181)***	0.594 (0.087)***	0.590 (0.087)***
District magnitude × open ballot		−0.378 (0.326)		−1.446 (0.688)*		−0.293 (0.313)
Constant	−5.653 (1.791)**	−5.828 (1.772)**	−2.225 (1.968)	−0.752 (1.915)	−4.978 (1.256)***	−5.093 (1.253)***
N (observations)	372	372	133	133	505	505
N (terms)	58	58	29	29	87	87
AIC	884	884	639	637	1521	1522
Log likelihood	−429	−428.4	−306.9	−304.8	−747.5	−747.1
SD (country-term)	0.64	0.61	0.57	0.42	0.76	0.75
Wald χ²	64.8***	68.2***	42.7***	49.7***	113.8***	116***

Note: Random effect negative binomial models. Dependent variable is the number of switches that a party witnesses in a given parliamentary term. ENP = effective number of parties; SD = standard deviation; AIC: Akaike information criterion.

*p < 0.05; **p < 0.025; ***p < 0.001.

Unlike vote- and office-seeking incentives, policy motives do drive politicians to switch in advanced democracies. As indicated by the positive and statistically significant estimate, a change in a party’s policy position increases the likelihood of switching. A change of one point in party policy increases switching by 0.03. If there is no shift in policy position (approximately one standard deviation below the mean), the party is predicted to witness 0.3 switches. However, a policy change of 22.4 points (one standard deviation above the mean) leads to 0.5 switches. In the case of the largest policy change (65.3 points), the party is expected to witness 1.1 switches, which is 3.5 times more than parties that have not changed policy position. The results suggest that party switching in advanced democracies is associated with policy rather than with legislators’ pursuit of re-election and office benefits. These findings conform to my expectations with respect to advanced democracies. In such polities, higher ideological transaction costs undermine the impact of vote- and office-seeking incentives on switching. However, policy-driven switching is unlikely to implicate such costs as the initial motivation of those switchers was a change in their party’s policy position.

The variable electoral volatility is positive and statistically significant, indicating that high levels of electoral volatility in advanced polities reduce the parties’ ability to keep their members from switching. Party switching is also affected by party size: larger parties tend to witness more switching. However, electoral rules, the effective number of parties¹⁴ and the duration of the term do not affect party switching. The interactive term of the two electoral variables in model 2 is statistically insignificant, indicating that the incentives for cultivating personal vote do not affect switching.¹⁵

Model 3 presents the negative binomial regression results for the effects of the motivational predictors on party switching in post-communist democracies in Europe. Both the variable vote share and the variable government are negative and statistically significant, indicating that losing parties and opposition parties are likely to witness switching. When a party anticipates getting in the next election only 35% of the vote share (one standard deviation below the mean) achieved in the last elections, 6.1 MPs are predicted to switch. When a party expects to increase its vote share by 185% (one standard deviation above the mean), only 4.6 switches are predicted, that is, 25% less. Opposition parties are likely to witness 6 switches during a parliamentary term, which is 30% more than government parties (4.6 switches). The results suggest that, unlike parties in advanced European democracies, parties in post-communist countries that cannot provide their MPs with re-election prospects or office benefits are less capable of preventing defection. However, a change in party policy does not affect party switching, as indicated by the statistically insignificant result for the variable policy.

The findings confirm the study’s hypotheses. The programmatic party competition, which is an attribute of European advanced democracies, renders vote- and office-driven switching costly. In other words, these defection costs deter mainly those politicians who seek to improve their re-election prospects and to maximize their office benefits by changing party membership. However, when a party changes its ideological ideal point, it may no longer be ideologically compatible with the members who remain in the party. Thus, in certain cases, changing party labels can even reward switchers ideologically.

Party switching in new polities is less affected by party labels than party switching in advanced democracies: politicians can advance their interests and switch parties without facing ideological costs, resulting in widespread vote- and office-driven switching. However, changes in parties’ policy position do not lead to party switching in systems where parties are less defined by clear ideological labels.

The personal vote interactive term presented in model 4 is negative and statistically significant, indicating that electoral systems that encourage MPs to cultivate their personal vote for their re-election are less likely to switch. District magnitude dampens switching in countries with an open ballot structure and does not have an impact on switching when the ballot structure is closed.¹⁶ Those findings are in line with previous studies on electoral systems’ effects on switching that argue that a candidate-centred environment reduces incentives to switch. In electoral systems that encourage the cultivation of candidate’s personal vote, parties have less influence on politicians’ re-election prospects. Hence, politicians will be less keen to defect from losing parties in order to improve their re-election chances, and they will focus more on building a personal vote coalition. In addition, party-centred systems increase switching as party lists can protect switchers from a possible electoral punishment by voters.

The electoral rule effects on switching can be found in post-communist democracies but not in advanced democracies. These findings support my main argument. In new democracies, the great utility from switching in order to improve re-election chances increases switching considerably. As electoral systems can increase or reduce parties’ capacity to determine MPs’ electoral fortune, it is reasonable that electoral systems would affect switching mainly in post-communist democracies, where switching is vote-driven.

Model 5 presents the results for all 25 countries included in the analysis and demonstrates that the variable perceptual disagreement is positive and statistically significant. This positive value indicates that higher levels of voter perceptual disagreement increase party switching. If the average perceptual disagreement in a given parliamentary term is 1.94 (one standard deviation above the mean), the predicted number of switches is 0.92, almost twice as much as in systems where the perceptual disagreement is lower (0.5 switches at one standard deviation below the mean). Thus, the finding confirms my expectation and bolsters the argument that the higher the degree of programmatic structuring of a party system, the higher the defection-deterring costs, which, in turn, reduce switching.

To further assess the causal relationship, I examine the ideological structuring effects on switching motives, dividing the data into two models along the perceptual disagreement measure (model 7–10 in the Online Appendix¹⁷). As hypotheses 1 and 2 indicate, I expect parties that cannot deliver policy benefits to witness switching mainly in advanced democracies with programmatically structured party systems. However, I also expect parties with the ability to provide their MPs with office and electoral benefits to impede switching mainly in systems with high levels of perceptual disagreement. The findings lend additional support to the hypotheses: ideological structuring of the party system increases parties’ ability to keep their MPs from seeking electoral and office – but not policy – benefits in other parties. Only in systems with low levels of perceptual disagreement do losing parties and winning parties face the same probability of witnessing switching. However, only in systems with limited perceptual disagreement are parties that remained ideologically stable more likely to retain their MPs than parties that have made a considerable shift.

To provide additional evidence for the mechanisms, I examine the determinants for switching (models 11–12 in the Online Appendix) in new democracies, that is, post-communist countries, in their early parliamentary terms and later terms. In the early parliamentary terms (first to third terms), party switching can be explained only by the ability of parties to secure re-election for their MPs. In later terms, when party competition becomes more ideologically structured, party switching can be explained by parties’ ability to deliver policy pay-offs – in addition to electoral incentives. The findings provide additional support for the argument that policy-driven switching can be found mainly in democracies with ideologically crystallized party systems. In other words, politicians would switch parties to fulfil their policy goals once the net utility from switching would be beneficial.

The variable open ballot in the later terms of the new democracies (model 11) is negative and statistically significant, indicating that candidate-centred systems reduce switching. Yet ballot structure does not affect switching in the earlier terms of the new democracies (model 12). The explanation lies in the notion of information and the establishment of voter–elite linkages. On the one hand, switchers in new democracies are motivated by re-election both in earlier and later terms. However, it is only in the later terms that voters, as well as politicians, have become familiar with and knowledgeable about the electoral system and its effects, and the strategic coordination role of the electoral institutions is established.

Explaining electoral volatility effects

As models 3–4 demonstrate, electoral volatility does not have a significant impact on switching in post-communist countries. This should be examined in light of the comparison with the positive and statistically significant results of the variable in advanced democracies (models 1–2). These findings are intriguing because they offer a novel insight into the ongoing debate on the relationship between a volatile electorate and erratic politicians. Recall that the literature acknowledges the differences between new and established democracies with respect to volatile voters. While electoral volatility in new polities is attributed to inexperienced voters who have not yet developed loyalty to parties (Tavits, 2008), electoral volatility in advanced democracies is understood as the behaviour of sophisticated voters who make an informed decision to switch parties (Van der Meer et al., 2015). The findings suggest that emancipated voters in institutionalized systems have the ability to trigger legislative party switching. By contrast, volatile elites in post-communist polities cannot be explained by instability in the electorate.¹⁸ The results are in line with the main argument of this study that switching in advanced countries is policy-driven. High levels of electoral volatility in advanced polities may indicate a shift in the ideological identity of large swaths of the electorate and can motivate representatives to react to this shift.

Conclusion

It is a well-established fact that parties that cannot provide their MPs with electoral, office and policy benefits are likely to experience switching. But are switchers in established democracies and new democracies driven by the same benefits? The findings demonstrate that while parties in new democracies tend to witness mainly vote- and office-driven switchers, MPs in advanced democracies are likely to switch on policy grounds. These findings are in accord with the main argument of this article that the ideological costs undermine the impact of vote- and office-seeking incentives on switching. In programmatically structured party systems, where the utility of holding office is a function of ideology, parties possess clear ideological labels, and switchers are likely to find themselves in ideologically incompatible parties, facing ideological costs for not being able to promote their agenda, and electoral costs for upsetting their party’s voters.

The article contributes to our understanding of party system stability and its consequences for democratic representation. The article proposes that ideological crystallization of party competition enhances parties’ ability to retain their MPs, even if the parties cannot offer electoral or office benefits. However, parties that change agenda during a parliamentary term are likely to witness switching on policy grounds.

An additional finding that bolsters the article’s argument reveals that electoral volatility in advanced democracies leads to volatile elites. These results do not hold in new democracies where fluctuation among elites is not a response to erratic voters. The findings reduce the gap in the relationship between electoral volatility and stability in the party system and highlight how important a programmatic principal–agent relationship is for the stability of party systems. Party switching should be seen as an indicator of weakly institutionalized party systems in new democracies where parties that fail to deliver vote and office pay-offs tend to witness defection. By contrast, party switching in advanced democracies should be understood as policy-driven and as the result of an ideological shift in the electorate. As legislators are responsive to their constituencies, they will be reluctant to defect from the party under whose label they have been elected – unless they are motivated by a shift in the ideological identity of the electorate, which might secure policy pay-offs for them. Hence, the article shows that ideological structuration of party competition renders parties more representative. While party switching is often perceived as a failure of democratic representation, politicians who switch parties on policy grounds in response to their constituencies’ ideology should be seen as representatives who accommodate rather than distort their voters’ will.

Supplemental material

Online_Appendix - Explaining legislative party switching in advanced and new democracies

Online_Appendix for Explaining legislative party switching in advanced and new democracies by Elad Klein in Party Politics

Footnotes

Acknowledgements

The author would like to thank Thomas M Meyer, Marcelo Jenny, the participants of the seminar series in the Department of Government at the University of Vienna, the Editor of Party Politics Paul Webb and three anonymous reviewers for their helpful suggestions.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Aarhus University Research Foundation (AUFF), ref. AUFF-E-2017-7-13.

ORCID iD

Elad Klein

Supplemental material

Supplemental material for this article is available online.

Notes

References

Albright

(2010) The multidimensional nature of party competition. Party Politics 16(6): 699–719.

Aldrich

(1995) Why Parties? Chicago: University of Chicago Press.

Aldrich

Bianco

(1992) A game-theoretic model of party affiliation of candidates and office holders. Mathematical and Computer Modelling 16(8–9): 103–116.

André

Depauw

Deschouwer

(2012) Legislators’ local roots: disentangling the effect of district magnitude. Party Politics 20: 904–917.

André

Depauw

Martin

(2015) Electoral systems and legislators’ constituency effort: the mediating effect of electoral vulnerability. Comparative Political Studies 48(4): 464–496.

Benoit

Laver

Mikhaylov

(2009) Treating words as data with error: uncertainty in text statements of policy positions. American Journal of Political Science 53(2): 495–513.

Butler

Stokes

(1974) Political Change in Britain. London: Macmillan.

Carey

Shugart

(1995) Incentives to cultivate a personal vote: a rank ordering of electoral formulas. Electoral Studies 14(4): 417–439.

Cox

(1997) Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press.

10.

Cox

McCubbins

(2005) Setting the Agenda: Responsible Party Government in the US House of Representatives. Cambridge: Cambridge University Press.

11.

Cox

McCubbins

(2007) Legislative Leviathan: Party Government in the House. Cambridge: Cambridge University Press.

12.

Dahl

(1971) Polyarchy: Participation and Opposition. New Haven: Yale University Press.

13.

Desposato

Scheiner

(2008) Governmental centralization and party affiliation: legislator strategies in Brazil and Japan. American Political Science Review 102(04): 509–524.

14.

Desposato

(2006) Parties for rent? Ambition, ideology, and party switching in Brazil’s chamber of deputies. American Journal of Political Science 50(1): 62–80.

15.

Dinas

Gemenis

(2010) Measuring parties’ ideological positions with manifesto data: a critical evaluation of the competing methods. Party Politics 16(4): 427–450.

16.

Downs

(1957) An Economic Theory of Democracy. New York: Harper and Row.

17.

Farrell

(2011) Electoral Systems: A Comparative Introduction. London: Macmillan International Higher Education.

18.

Franzmann

Kaiser

(2006) Locating political parties in policy space: a reanalysis of party manifesto data. Party Politics 12(2): 163–188.

19.

Gemenis

(2012) Proxy documents as a source of measurement error in the Comparative Manifestos Project. Electoral Studies 31(3): 594–604.

20.

Gemenis

(2013) What to do (and not to do) with the Comparative Manifestos Project data. Political Studies 61(suppl 1): 3–23.

21.

Heller

Mershon

(2005) Party switching in the Italian chamber of deputies, 1996–2001. Journal of Politics 67(2): 536–559.

22.

Heller

Mershon

(2008) Dealing in discipline: party switching and legislative voting in the Italian chamber of deputies, 1988–2000. American Journal of Political Science 52(4): 910–925.

23.

Herron

(2002) Causes and consequences of fluid faction membership in Ukraine. Europe-Asia Studies 54(4): 625–639.

24.

Kato

(1998) When the party breaks up: exit and voice among Japanese legislators. The American Political Science Review 92(4): 857–870.

25.

Kato

Yamamoto

(2009) Competition for power: party switching and party system change in Japan. In: Heller

Mershon

(eds) Political Parties and Legislative Party Switching. New York: Palgrave Macmillan, pp. 233–263.

26.

Kitschelt

(2000) Linkages between citizens and politicians in democratic polities. Comparative Political Studies 33(6–7): 845–879.

27.

Kitschelt

Hawkins

Luna

, et al. (2010). Latin American Party Systems, Cambridge Studies in Comparative Politics. New York: Cambridge University Press.

28.

Kitschelt

Mansfeldova

Markowski

, et al. (1999) Post-Communist Party Systems: Competition, Representation, and Inter-Party Cooperation. Cambridge: Cambridge University Press.

29.

Klein

(2016) Electoral rules and party switching: how legislators prioritize their goals. Legislative Studies Quarterly 41(3): 715–738.

30.

Klein

(2018) The personal vote and legislative party switching. Party Politics 24(5): 501–510.

31.

Kreuzer

Pettai

(2003) Patterns of political instability: affiliation patterns of politicians and voters in post-communist Estonia, Latvia, and Lithuania. Studies in Comparative International Development 38(2): 76–98.

32.

Laver

Benoit

(2003) The evolution of party systems between elections. American Journal of Political Science 47(2): 215–233.

33.

McElroy

(2003) Party switching in the European Parliament: why bother? Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago.

34.

McLaughlin

(2011) Electoral regimes and party-switching: floor-crossing in South Africa’s local legislatures. Party Politics 18(4): 563–579.

35.

McMenamin

Gwiazda

(2011) Three roads to institutionalisation: vote-, office- and policy-seeking explanations of party switching in Poland. European Journal of Political Research 50(6): 838–866.

36.

Mainwaring

Scully

(1995) Building Democratic Institutions: Party Systems in Latin America. Stanford: Stanford University Press.

37.

Mainwaring

Torcal

(2006) Party system institutionalization and party system theory after the third wave of democratization. In: Katz

Crotty

(eds) Handbook of Political Parties. London: Sage, pp. 204–227.

38.

Mair

(1997) Party System Change: Approaches and Interpretations. Cambridge: Cambridge University Press.

39.

Mershon

Shvetsova

(2013a) The microfoundations of party system stability in legislatures. The Journal of Politics 75(04): 865–878.

40.

Mershon

Shvetsova

(2013b) Party System Change in Legislatures Worldwide. Cambridge: Cambridge University Press.

41.

Müller

Strøm

(1999) Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press.

42.

Nielsen

Andersen

Pedersen

(2018) Balancing costs of legislative party switching in the Danish Parliament 1953–2015. Parliamentary Affairs 72(1): 42–58.

43.

Nokken

(2009) Party switching and the procedural party agenda in the U.S. House of Representatives. In: Heller

Mershon

(eds) Political Parties and Legislative Party Switching. New York: Palgrave Macmillan pp. 81–108.

44.

Nokken

Poole

(2004) Congressional party defection in American history. Legislative Studies Quarterly 29(4): 545–568.

45.

Norris

(1997) Passages to Power: Legislative Recruitment in Advanced Democracies. Cambridge: Cambridge University Press.

46.

O’Brien

Shomer

(2013) A cross-national analysis of party switching. Legislative Studies Quarterly 38(1): 111–141.

47.

Pedersen

(1983) Changing patterns of electoral volatility in European party systems, 1948–1977: explorations in explanation. In: Daalder

Mair

(eds) Western European Party Systems: Continuity and Change. London: SAGE, pp. 29–66.

48.

Powell

Tucker

(2014) Revisiting electoral volatility in post-communist countries: new data, new results and new approaches. British Journal of Political Science 44(01): 123–147.

49.

Przeworski

(1975) Institutionalization of voting patterns, or is mobilization the source of decay? American Political Science Review 69(01): 49–67.

50.

Reed

Scheiner

(2003) Electoral incentives and policy preferences: mixed motives behind party defections in Japan. British Journal of Political Science 33(03): 469–490.

51.

Rose

(1995) Mobilizing demobilized voters in post-communist societies. Party Politics 1(4): 549–563.

52.

Rose

Munro

White

(2001) Voting in a floating party system: the 1999 Duma election. Europe-Asia Studies 53(3): 419–443.

53.

Schumpeter

([1942] 2013) Capitalism, Socialism and Democracy. New York: Routledge.

54.

Sevi

Yoshinaka

Blais

(2018) Legislative party switching and the changing nature of the Canadian party system, 1867–2015. Canadian Journal of Political Science 51(3): 665–695.

55.

Shabad

Slomczynski

(2004) Inter-party mobility among parliamentary candidates in post-communist East Central Europe. Party Politics 10(2): 151–176.

56.

Taagepera

Grofman

(2003) Mapping the indices of seats–votes disproportionality and inter-election volatility. Party Politics 9(6): 659–677.

57.

Tavits

(2005) The development of stable party support: electoral dynamics in post-communist Europe. American Journal of Political Science 49(2): 283–298.

58.

Tavits

(2008) On the linkage between electoral volatility and party system instability in Central and Eastern Europe. European Journal of Political Research 47(5): 537–555.

59.

Tavits

Letki

(2009) When left is right: party ideology and policy in post-communist Europe. American Political Science Review 103(04): 555–569.

60.

Thames

(2016) Electoral rules and legislative parties in the Ukrainian Rada. Legislative Studies Quarterly 41(1): 35–59.

61.

Thames

(2007) Searching for the electoral connection: parliamentary party switching in the Ukrainian Rada, 1998–2002. Legislative Studies Quarterly 32(2): 223–256.

62.

Van der Brug

(2003) Analysing party dynamics by taking partially overlapping snapshots. In: Laver

(ed) Estimating the Policy Position of Political Actors. New York: Routledge, pp. 135–152.

63.

Van der Brug

Van Spanje

(2009) Immigration, Europe and the ‘new’ cultural dimension. European Journal of Political Research 48(3): 309–334.

64.

Van der Meer

Van Elsas

Lubbe

, et al. (2015) Are volatile voters erratic, whimsical or seriously picky? A panel study of 58 waves into the nature of electoral volatility (The Netherlands 2006–2010). Party Politics 21(1): 100–114.

65.

Volkens

Lehmann

Merz

, et al. (2014) The Manifesto Data Collection. Manifesto Project (MRG/CMP/MARPOR). Version 2014b. Berlin: Wissenschaftszentrum Berlin für Sozialforschung (WZB).

66.

Young

(2014) An initial look into party switching in Africa: evidence from Malawi. Party Politics 20(1): 105–115.

67.

Zaller

(1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.20 MB