Opposition goals and government support

Various explanations of opposition behaviour can be brought together in a single theoretical framework using the classic policy/office/votes triad of party behaviour (Strøm Reference Strøm1990a; Strøm and Müller Reference Strøm, Müller, In Müller and Strøm1999). In this framework, opposition parties are expected to choose the course of action that they believe, under the present institutional constraints, will have the best consequences for their preferences. This framework distinguishes between three distinct preferences or goals that parties aim to achieve: maximum electoral support (votes), control of cabinet positions (office) and changing public policies (policy).

The first two goals are easily identified in the stereotypical image of government–opposition relations shaped by Prime Minister’s Questions in the British House of Commons. The leader of the opposition arguing strongly with the prime minister in parliament reflects the adage of 19th-century statesman Edward Stanley, the Earl of Derby, that ‘The Duty of an Opposition [is] very simple … to oppose everything, and propose nothing!’ The Westminster depiction of a parliamentary opposition is an opposition with barely any leverage against the government, besides ‘good reasons, and time’ (King Reference King1976: 18). In the short term, then, the opposition prioritizes votes over policy and office: it is a ‘government-in-waiting’ that can only aim to win the next elections to replace the incumbent government and obtain control over government policies (Hix and Noury Reference Hix and Noury2016; Laver Reference Laver, In Wittman and Weingast2006; Shugart Reference Shugart, Binder, Rhodes and Rockman2006). Though the relationship between electoral results, government participation and policy influence is less direct in multiparty systems, even here votes matter greatly to parties for them to influence policies or to increase the likelihood of coalition participation (Downs Reference Downs1957: 159).

Opposition behaviour in parliament influences electoral outcomes. Voters evaluate the performance of governing parties more negatively when government actions are criticized and opposed by opposition parties (Seeberg Reference Seeberg2020). Whereas voters can hold governing parties accountable for their performance in office, they can hold opposition parties accountable for their behaviour in parliament. Opposition parties are generally assumed to benefit electorally from criticizing and opposing the government. Opposition parties who cooperate structurally with minority governments are more disliked by voters who are dissatisfied with the government and they generally lose votes in the next elections (Hjermitslev Reference Hjermitslev2024; Thürk and Klüver Reference Thürk and Klüver2024). Although many voters detest unconstructive and disruptive opposition (Louwerse and Zorina Reference Louwerse and Zorina2025; Stiers Reference Stiers2023), opposition parties that vote more frequently against the government in parliament generally do better in the next elections than opposition parties that vote more frequently with the government (Tuttnauer and Wegmann Reference Tuttnauer and Wegmann2022; Williams Reference Williams2011). Opposing the government meets people’s expectations of a good opposition, and it may help opposition parties present themselves as an ideologically distinct alternative to the incumbent government. When voters see no differences between the government and the opposition, they will feel no need to replace the government with the opposition.

As long as voters do not notice parliamentary voting behaviour, opposition parties might get away with supporting the government. To explain opposition parties’ behaviour, however, it matters to what degree parties think that their votes in parliament shape the public’s perception of them. Voting on government bills is one of the most important and visible activities in parliament. Considering that parliamentary behaviour shapes media narratives about interparty conflicts (Fortunato Reference Fortunato2021) and parties often advertise their and each other’s voting decisions on social media and press releases (Tuttnauer and Wegmann Reference Tuttnauer and Wegmann2022), it is easy to understand how opposition parties vote strategically to appeal to voters (Louwerse and Otjes Reference Louwerse and Otjes2019; Tuttnauer Reference Tuttnauer2018).

Voting against the government’s legislation can be particularly attractive as it is considered a ‘mediatized signalling tool for opposition parties as an explicit and clearly attributable expression of party positions and their attitude toward the government’ (Tuttnauer and Wegmann Reference Tuttnauer and Wegmann2022: 1357). Here, opposition parties have a strategic advantage in deciding their position, because they ‘usually do not have to commit themselves on any issue until after … the government has revealed its policy’ and then they ‘can select the optimum strategy to counteract it’ (Downs Reference Downs1957: 52). By opposing the government’s position, opposition parties send signals to voters who are unsatisfied with the incumbent government (Ganghof and Bräuninger Reference Ganghof and Bräuninger2006; Tuttnauer Reference Tuttnauer2018) (see Table 1).

Table 1. The Goals of Opposition Parties and Corresponding Behaviour in Parliament

Next to the desire to win elections, office aspirations can also influence opposition parties’ relations with the government. Here, seeking office means aiming to maximize a party’s share of cabinet positions. Opposition parties may pursue office because of the intrinsic value of having ministerial posts – ‘income, prestige, and power’ (Downs Reference Downs1957: 28) – or for its instrumental value: to influence public policies ‘it is almost always better to be in office than not’ (Strøm and Müller Reference Strøm, Müller, In Müller and Strøm1999: 6). Regardless of the reason for pursuing office, office-seeking means wanting to leave the opposition and become or join the government. Office-seeking motivations may lead opposition parties to oppose the government if that helps them to foster a government defeat in parliament or stimulate conflicts within the government, resulting in early elections or a government alternation. On the other hand, they may also lead opposition parties to support the government to signal trustworthiness and responsibility as potential partners for a future coalition government (Louwerse et al. Reference Louwerse2017). The best strategy for opposition parties to leave the opposition depends on how credible they can make a claim to replace the incumbent government after the next elections (Tuttnauer Reference Tuttnauer2018).

For opposition parties, vote-seeking and office-seeking behaviour are investments for the future – that is, the next government formation or the next elections. However, parties do not necessarily have to wait until it is their turn to govern to ‘maximize [their] impact on public policy’ in the short term (Strøm and Müller Reference Strøm, Müller, In Müller and Strøm1999: 7, emphasis added). Although policy-seeking behaviour is more associated with the government and government legislation, ‘even opposition parties can enjoy some policy influence in most parliamentary democracies’ and, therefore, the ‘role of the opposition is not simply to criticize and present an alternative government’ (Strøm Reference Strøm1990b: 42). Here, opposition parties can try to accomplish some policy goals, not only by introducing private members’ bills but also by shaping the parliamentary agenda, cooperating with policy specialists from the government parties and introducing amendments. The policy influence differential varies across parliaments (Garritzmann Reference Garritzmann2017; Helms Reference Helms2004; Powell Reference Powell2000; Strøm Reference Strøm1990b: 42; Wegmann Reference Wegmann2022). Still, rather than aiming to leave the opposition, the opposition in any parliament can opt to exploit its position to aim for policy gains (Ilonszki and De Giorgi Reference Ilonszki, De Giorgi, De Giorgi and Ilonszki2018: 5).

Initiating most legislation, the government remains the central policymaking body in most political systems. Instead of distinguishing themselves from the government, opposition parties may opt to cooperate with the government to change public policies in a more favourable direction. During the legislative review process, they can influence policies, particularly when parliamentary committees enable them to monitor and control government policymaking (Fortunato Reference Fortunato2019; König et al. Reference König, Lin and Silva2023; Strøm Reference Strøm1990b). Opposition parties may also negotiate policy concessions in exchange for their support (Anghel and Thürk Reference Anghel and Thürk2021; Christiansen and Seeberg Reference Christiansen and Seeberg2016; Otjes and Louwerse Reference Otjes and Louwerse2024).

Even when governments do not need the numerical support of opposition parties, the wishes of opposition parties are sometimes still accommodated by the government to ‘silence’ opposition criticism (Green-Pedersen and Mortensen Reference Green-Pedersen and Mortensen2010; Seeberg Reference Seeberg2013, Reference Seeberg2023; Thesen Reference Thesen2013). Although opposition parties may strongly attack and criticize the government to put new issues on the political agenda, opposition parties will ultimately support government bills that accommodate their wishes. Typically this adversarial rhetorical behaviour is also motivated by opposition parties’ efforts to ‘influence elections and voters’ evaluations of the government’ rather than efforts to change the government’s policy, even though this behaviour ‘may have policy implications because the government legislates to silence opposition agenda-setting’ (Seeberg Reference Seeberg2023: 466). Therefore, policy-seeking motivations may generally lead opposition parties to support the government, either because the government’s legislation is closer to opposition parties’ policy preferences than the status quo or because the government accommodates the wishes of opposition parties in exchange for support.

Opposition goals trade-offs and expectations

Even though winning elections, becoming the government and exerting policy influence are clearly interrelated objectives, it should be clear by now that also for parties in the opposition ‘[t]he same behaviour that maximizes one [objective] may not lead to the best possible outcome with respect to the others’ (Strøm and Müller Reference Strøm, Müller, In Müller and Strøm1999: 9). Depending on their ‘time horizons’, parties prioritize a certain goal over others, at least in the short term. Because parties in the opposition engage in different strategic behaviours to achieve office, policy and votes (Ilonszki and De Giorgi Reference Ilonszki, De Giorgi, De Giorgi and Ilonszki2018), making an analytical distinction between these three goals helps understand why opposition parties support or oppose the government.

Depending on the (institutional) context, opposition parties face different trade-offs between their goals, and they will exhibit different strategic behaviours in parliament. One important trade-off exists between policy-seeking and vote-seeking opposition behaviour (Andeweg Reference Andeweg, In Müller and Narud2013). Voters may become dissatisfied with opposition parties who continually support or work together with the government. Julian Garritzmann (Reference Garritzmann2017: 8) describes the tension between opposition roles: ‘political actors seeking to remain credible cannot shape policy-making and simultaneously criticize and present distinct alternatives to results that they themselves have shaped; other political actors would highlight the opposition’s inconsistent behaviour.’

A crucial factor that shapes the way opposition parties weigh policy-seeking and vote-seeking behaviour against each other is their policy positions (De Giorgi and Marangoni Reference De Giorgi and Marangoni2015; Hohendorf et al. Reference Hohendorf, Saalfeld, Sieberer, Bukow and Jun2020). It is very unattractive for an opposition party to support government policies when there is a large difference between the party’s and the government’s policy positions. Even if an opposition party manages to receive some policy concessions from the government, it is highly unlikely that the result will leave their voters satisfied in that case (Ganghof and Bräuninger Reference Ganghof and Bräuninger2006). If an opposition party’s position on a particular issue is close to the government’s, that opposition party may find it easier to support the government: the party may consider the government’s policy an improvement over the status quo or it may be easier to compromise with the government over such an issue without giving too much away. This will leave voters more satisfied than if opposition parties should support government policies that are further away from the party’s ideal point. For governing parties too, it is more attractive to seek cooperation with and accommodate ideologically close opposition parties than with ideologically distant opposition parties (Louwerse et al. Reference Louwerse2017):

The trade-off between policy-seeking and vote-seeking behaviour may also tilt further in favour of policy-seeking when the opposition possesses veto power over legislation – for example under minority governments. Substantive minority governments must build ad hoc majority support for every proposal in order to survive and implement its policies (Louwerse et al. Reference Louwerse2017; Strøm, 1990). Opposition parties may allow the survival of minority governments in order to obtain substantial policy gains (Artés and Bustos Reference Artés and Bustos2008). This only holds for ‘genuine’ minority governments, though. There are important differences between minority governments working on an issue-by-issue basis with different opposition parties (substantial minority governments) and those with long-term partnerships with formal support parties (Strøm Reference Strøm1990b; Thürk and Klüver Reference Thürk and Klüver2024). The latter case differs only a little from the situation of a majority government. These formal minority governments have secured majority support for their main policies and confidence votes and are ‘majority governments in disguise’ (Strøm Reference Strøm1990b: 61), evidenced by similar levels of legislative performance and survival rates to majority governments (Krauss and Thürk Reference Krauss and Thürk2022; Thürk Reference Thürk2022).

Under substantive minority governments, opposition parties are in a stronger position to change policies or prevent undesirable changes. These minority governments will try to accommodate opposition parties to safeguard majority support in parliament for their legislation. They negotiate support from opposition parties before or during the legislative review process (Christiansen and Damgaard Reference Christiansen and Damgaard2008; Christiansen and Pedersen Reference Christiansen and Pedersen2014). This means there are more incentives for opposition parties to opt for policy-seeking behaviour when making a trade-off between policy and votes. Consequently, opposition parties will be more likely to vote in favour of those government bills.

Additionally, the role of opposition parties’ policy preferences may be stronger under a substantive minority government than under a majority government. Under majority governments, opposition parties have stronger incentives to oppose the government even if they prefer government legislation over the status quo (Dewan and Spirling Reference Dewan and Spirling2011). Under substantive minority governments, opposition parties can tilt the support for a government bill to a majority. Thus, under minority governments, support for the government will be affected more by parties’ ‘policy positions rather than by government–opposition splits’ (Hix and Noury Reference Hix and Noury2016: 254; Klüver and Zubek Reference Klüver and Zubek2018).

Opposing the government to win votes at the next election can come with a trade-off of forgoing any possible policy influence while being in opposition and at a lower chance of joining a future coalition government with one of the incumbent governing parties. Opposition parties are expected to prioritize vote-seeking behaviour in situations where that behaviour will be most effective: when opposition parties feel more strongly monitored by voters, they will feel incentivized to do things that will benefit them electorally. When opposition parties ‘do their work away from the centers of public attention’ they are more inclined to cooperate with the government as ‘opportunities are more available and compromises less embarrassing’ (Strøm Reference Strøm1990b: 43).

Opposition parties should be expected to be monitored most closely by the electorate on their core issues. Their behaviour in their most important policy areas is more important for preserving a party profile distinct from the government (Budge Reference Budge2015). On salient issues on which they have very different preferences from the government, opposition parties should be particularly oppositional. Opposing the government is probably not equally beneficial electorally for all parties in all contexts. When opposition parties are close to the government on issues that are important to them, they may be more willing to support the government. It is harder for opposition parties to credibly oppose government legislation that is congruent with (or very close to) their policy preferences. This is especially true in multiparty systems in which opposition parties have to position themselves vis-à-vis other parties in the opposition in addition to the government (Hohendorf et al. Reference Hohendorf, Saalfeld, Sieberer, Bukow and Jun2020). Opposing the government in that situation is not only less credible, but gaining policy influence may be more likely and less costly because it is easier for parties to justify their support for the government and be rewarded for it by their voters (Müller Reference Müller2024; Thesen et al. Reference Thesen, Green-Pedersen and Mortensen2017).

Office-seeking motivations may lead opposition parties to oppose the government if that helps them advance an early government alternation. The threat of a parliamentary defeat may unite the governing party or parties in their support of the government and incentivize the opposition to vote collectively against the government (Hix and Noury Reference Hix and Noury2016: 252). Although most governments terminate early rather than serving their full term, they rarely terminate because of a defeat in parliament by the opposition. For coalition governments, the most frequent reason for early termination is a policy conflict between the coalition partners (Damgaard Reference Damgaard, Strøm, Müller and Bergman2008).

Bridging diverging policy preferences through compromise can be very painful for coalition parties. The constitutional doctrine of collective cabinet responsibility can be a risk for coalition parties. According to this doctrine, all members of the government should support all government actions. As a consequence, coalition parties in parliament generally vote in the same way on government legislation, even if they have very diverging policy preferences (Laver Reference Laver, In Wittman and Weingast2006: 137). They will not allow their partners to vote against the government compromise. Coalition unity may hurt them electorally because voters may feel that their party is giving too much away (Hjermitslev Reference Hjermitslev2020; Klüver and Spoon Reference Klüver and Spoon2020). For coalition parties, collective cabinet responsibility is ‘an electoral straitjacket, effectively eliminating a cabinet member’s potential to defend themselves’ (Fortunato Reference Fortunato2021: 94).

Opposition parties that want to leave the opposition or prefer a different government can exploit coalition parties’ electoral vulnerability by increasing intra-coalition tensions. An effective way to do this is to expose and highlight disagreements or ‘policy wedges’ between the governing parties (Van de Wardt et al. Reference Van de Wardt, De Vries and Hobolt2014). Exposing policy differences and compromises among coalition parties can destabilize a coalition government (Laver Reference Laver, In Wittman and Weingast2006: 130). It embarrasses coalition parties and communicates to voters that coalition parties have given away too much to their partners (Fortunato Reference Fortunato2021). This will hurt coalition parties electorally at the next elections, and it may result in increasing tensions among coalition parties and in the early termination of the government (So Reference So2024).

One way to ‘divide to conquer’ the government is by asking parliamentary questions on policy issues on which the governing parties’ policy preferences diverge (Whitaker and Martin Reference Whitaker and Martin2022). Another way of publicizing these policy wedges is voting against government bills that concern policy domains on which coalition parties diverge ideologically. A situation in which a governing party tilts the vote to a majority for unpopular policies may be especially painful for it. This strategy may not only hurt the stability and prospects of government duration but also hurt the governing parties’ reputations and electoral prospects at the next election.

The prospects of a party that uses an adversarial opposition style entering the next government depends on the governing potential of opposition parties and the government alternation patterns in a political system. Governmental potential refers to opposition parties’ aspirations for office and the likelihood of getting involved in government formation and is often operationalized as whether that party has had prior government experience (Sartori Reference Sartori1976; Schumacher et al. Reference Schumacher2015; Tuttnauer Reference Tuttnauer2018). Opposition parties with none or only little governing potential will probably put a stronger weight on policy-seeking or vote-seeking motivations than on office-seeking motivations. The strategy for potential governing parties to win office depends on the structure of party competition (Mair Reference Mair, LeDuc and Niemi1996; Otjes Reference Otjes2020).

In systems of wholesale alternation, two parties or two blocs of parties alternate in government. The route to government for office-seeking opposition parties in these systems is to win a majority (or plurality) of seats (or for their bloc) at the next election. In these systems, early elections and being distinct from the government are the most rational routes to becoming the government in the long term. Attacking the government, then, is a requirement to offer a credible alternative government to the electorate (Tuttnauer Reference Tuttnauer2018).Footnote ¹ For these opposition parties, ‘the basic conviction … prevails, “that it is better to give the government enough rope to hang itself with” than demonstrating its fitness for office and trying to win the voters’ trust by helping the government to govern’ (Helms Reference Helms2008: 15). For opposition parties without governing potential, campaigning against the government is not a self-evident strategy. As they are unlikely to gain government participation through winning votes, their only way to influence policies is by exploiting the parliamentary opportunities to influence public policies. To do that, they need to act more cooperatively towards the government (Tuttnauer Reference Tuttnauer2018).

Contrarily, in systems of partial alternation, parties that are separated into government and opposition camps at one time may form a coalition government together at another time. Elections in these systems – however important – are less deterministic for government formation. Here, office-seeking opposition parties risk being excluded from future coalitions when they act too antagonistically towards the government (Ganghof and Bräuninger Reference Ganghof and Bräuninger2006; Louwerse et al. Reference Louwerse2017; Tavits Reference Tavits2008). Therefore, they will moderate their attitude towards the government or even actively seek cooperation, signalling trust and gaining a reputation as future coalition partners (Bale and Bergman Reference Bale and Bergman2006). Cooperative behaviour may affect parties’ reputation positively as reliable partners in future coalition governments. Compared to permanent opposition parties, engaging in ‘divide to conquer’ strategies could hurt opposition parties with governing potential in systems of partial alternation because this ‘could risk imperilling relationships with past and prospective coalition partners’ (Van de Wardt et al. Reference Van de Wardt, De Vries and Hobolt2014: 987).

Methods and data

The analyses include parliamentary votes in four established parliamentary democracies: Canada, Denmark, the Netherlands and the United Kingdom. These countries were selected to maximize institutional variation and increase the findings’ generalizability. To inform the country case selection, two institutional features known to affect government–opposition distinctiveness were considered. First, the difference between two-party systems (the United Kingdom and Canada) and multiparty systems (the Netherlands and Denmark) represents differences in the competition between political parties (Lijphart Reference Lijphart2012). The second feature that informed the country case selection is the occurrence of minority governments. Denmark has the largest share of minority governments of all established parliamentary democracies: roughly 90% of Danish governments since 1945 had a minority status. Compared to Denmark, Canada has had far fewer minority governments, but for a majoritarian democracy, a third of the governments having had a minority status is still quite substantial. In the United Kingdom and the Netherlands, majority governments have been the norm. This diverse set of country cases contributes to the generalizability of the findings.

Additionally, the selected countries differ in their patterns of government alternation (Otjes Reference Otjes2020). In the United Kingdom, the Conservative Party and the Labour Party have alternated in mostly single-party government. In Canada, the Liberal Party and the Conservative Party have competed over single-party governments. Despite the Danish parliament’s high levels of fragmentation, the party system can be divided into a left bloc (Social Democrats and Social Liberals) and a right bloc (Liberals and the Conservative People’s Party) that usually alternate in single-party or coalition government. Deviating from these patterns of wholesale alternation, the Dutch fragmented party systems have led to the formation of coalition governments including at least one party that had previously been in government. In the Netherlands, there has never been an instance of wholesale alternation and, until the formation of the 2024 government, every government had included at least two of the three traditional parties: the Christian Democrats (or its predecessors), the Liberals and the Labour Party.

This study focuses on the decision of opposition parties to support or oppose government legislation. Whereas earlier research focused on the differences in government support between opposition parties (De Giorgi and Marangoni Reference De Giorgi and Marangoni2015; Tuttnauer Reference Tuttnauer2018) or on differences in the government–opposition distinction between parliamentary votes (Giuliani Reference Giuliani2008; Louwerse et al. Reference Louwerse2017), this analysis focused on the individual voting decisions of opposition parties, enabling the estimation of the impact of party- and bill-specific features on opposition behaviour (cf. Müller Reference Müller2024).

The unit of analysis is an opposition party’s voting decision on a government bill (in favour or against). Data on the voting behaviour in parliament of opposition parties were taken from Tom Louwerse and Rick van Well (Reference Louwerse and van Well2023), who assembled parliamentary voting data from various country-specific sources. This dataset spans an extensive period, from approximately 1945 to 2020 in all four parliaments. From this dataset, all voting decisions of opposition parties on government bills from non-caretaker governments were selected. Abstentions and parliamentary votes in which governing parties voted against the bill were excluded. The data on votes in the British House of Commons did not allow a distinction between government and private members’ bills and therefore all bills that the government party voted in favour of were included. Most of these will be government bills as it is rare for votes to take place on private members’ bills. Finally, votes on bills with missing information on bill topics or party were excluded ultimately resulting in 76,243 voting decisions on 14,148 government bills made by 351 opposition parties in 83 cabinet terms in 4 countries (see Table 2).

Table 2. Overview of Empirical Observations

The dependent variable in this study is the opposition parties’ support for government bills. Opposition parties’ support for government legislation was captured in a simple dichotomous measure: whether the majority of a party voted in support of (1) or in opposition to (0) the government bill. It must be noted that the meaning of an opposition party’s voting decision may differ slightly across parliaments, depending on the voting procedure. In the Danish Folketing, all government bills get a vote, whereas in the British House of Commons probably only the most politically salient bills get to a vote because of the cumbersome voting procedure (‘division’) (Hug Reference Hug, In Narud and Müller2013). Since the number of bills is quite different across parliaments, the degree of government–opposition distinctiveness cannot be truly compared between countries. Many of the bills voted on in the Netherlands and Denmark may be less controversial than the typical bill voted on in the United Kingdom or Canada. For the analyses conducted here, the limitations of this operationalization are not too problematic, though, as they focus mainly on explaining within-country variation. Furthermore, a study of voting in the German parliament found that ‘opposition behavior does not differ systematically between recorded and nonrecorded votes suggesting that parties act consistently across all votes’ (Hohendorf et al. Reference Hohendorf, Sieberer and Wenzig2024). The sample of bills that were voted on in parliament is therefore likely to be a valid measure for studying opposition behaviour in parliament.

For the operationalization of policy distance, issue saliency and coalition issue divisiveness, information on the topic of the bill was required. This information was largely retrieved from the Comparative Agendas Project (CAP). This project classifies various policy activities, including bills, into a single, universal and consistent coding scheme (Baumgartner et al. Reference Baumgartner, Breunig and Grossman2019). The coding scheme includes 21 major topics and over 200 subtopics used to hand-code policy activities. The bill topics for Denmark (1953–2017), Canada (1960–2010) and the Netherlands (1980–2009) provided by these data resources were matched to the parliamentary voting data.Footnote ² The Dutch data were extended to 2021 by a coder using the Dutch master codebook. Because the British voting data and CAP data did not share a similar unique identifier and the number of votes on British government bills was limited, these were also hand-coded using the British master codebook.

The policy preferences of political parties were measured using data from the Manifesto Research on Political Representation (MARPOR) dataset (previously known as the Comparative Manifestos Project (CMP)). MARPOR has collected and hand-coded election manifestos since 1945 in over 50 countries, including the four country cases in this study (Volkens et al. Reference Volkens2020). Sentences in parties’ election manifestos were sorted into one of 56 categories, resulting in percentages of each manifesto falling into each category. Rather than relying on the general left–right party positions that can be derived from these data, a more fine-grained measurement of party positions on specific policy dimensions was used; seven policy areas were constructed for which MARPOR categories were matched to CAP policy areas (see Appendix A in the Supplementary Material). For the estimation of issue-specific party, Will Lowe et al.’s (Reference Lowe2011) scaling method was used, taking the logged ratio of positive and negative statements in each category.

To estimate the policy distance between opposition parties and the government, the government’s position on each policy issue was calculated first. For single-party governments, the manifesto measures of the governing party were assumed to represent the government’s positions. The policy positions of coalition governments were operationalized as the seat-weighted average of the policy positions of the coalition parties, whereby formal support parties to minority governments were treated as coalition parties. The policy distances between parties and the government were calculated by taking the absolute distance between a party’s position and the government’s position. Coalition issue divisiveness was operationalized as the range – that is, the difference between the highest and lowest value – of coalition parties’ policy positions. For the estimation of issue saliency, Lowe et al.’s (Reference Lowe2011) method was used, taking the logged proportion of statements, whether positive or negative, in each category.

Information about cabinet type was derived from ParlGov (Döring and Manow Reference Döring and Manow2021). Formal support parties of minority governments were identified using Torbjörn Bergman et al.’s study (Reference Bergman, Bäck and Hellström2021). Minority governments that received majority support together with their formal support parties were classified as formal minority governments and other minority governments as substantive minority governments. Dummies for substantive minority government, as well as formal minority government and surplus majority government, were included in the analyses. As a proxy for governing potential, a continuous measure of government experience was used by dividing the number of days an opposition party had been in government by the total number of days this party had existed since 1918 (Schumacher et al. Reference Schumacher2015; Tuttnauer Reference Tuttnauer2018). As control variables, seat share (based on ParlGov), a formal support party dummy and policy area were included. See Appendix B in the Supplementary Material for descriptive statistics.

Findings

The data were structured in party–bill dyadics: voting decisions were nested into both opposition parties and bills, but parties and bills did not have a hierarchical structure. As the data were clustered within cabinet terms and bills they were analysed with multilevel logistic regression models, including random intercepts for cabinet and bill. Party-fixed effects (and effectively country-fixed effects) were also added to the model. The results of the pooled models are presented in Table 3.

Table 3. Multilevel Logistic Regression Analysis of Opposition Support for Government Bills

Notes: In all models, the number of observations is 76,243, the number of bills are 76,243, and the number of cabinets is 83. The reference category for cabinet type is minimal winning.

*** p < 0.001; ** p < 0.01; * p < 0.05; ^† p < 0.1

The analyses offer empirical support for the policy distance hypothesis. The larger the differences between the positions of an opposition party and the government on a certain policy area, the lower the probability that the opposition party votes in favour of government bills in that area. This pattern of opposition behaviour was found across political systems and was statistically significant in all parliaments except for the Canadian House of Commons (see Appendix C, Table C1, in the Supplementary Material). On average, the probability of voting for a government bill decreases by 1.3 percentage points, with each standard deviation increase on the policy distance measure. This finding aligns with a ‘non-strategic’ spatial model of parliamentary voting behaviour in which opposition parties decide to support or oppose government bills based purely on the substantive content of the policies proposed (Hohendorf et al. Reference Hohendorf, Saalfeld, Sieberer, Bukow and Jun2020).

It also fits, however, the expectation that opposition parties make a trade-off between policy gains through government support and winning votes through opposing the government. When the positions of an opposition party and the government are too far apart, it is harder to credibly support the government. Supporting the government on policy issues on which an opposition party’s initial position is closer makes it easier to find compromises that are not too painful for opposition parties. Still, the size of the policy distance effect is modest, confirming again that spatial models of legislative behaviour are not sufficient to understand opposition behaviour in parliament adequately. To provide evidence for the main argument in this article that opposition parties act strategically, balancing their opposition goals, I now turn to the other findings.

Opposition support for the government was expected to increase under substantive minority governments. Substantive minority governments should be more willing to accommodate opposition parties in exchange for their much-needed support in parliament. For opposition parties in turn this should increase the benefits of policy-seeking behaviour and tilt the policy–vote trade-off further in favour of policy-seeking behaviour – that is, support of government bills. The probability of voting for government bills among opposition parties to substantive minority governments is 5.3 percentage points greater, on average, than the probability of voting for government bills among opposition parties to minimal winning governments. The differences between opposition behaviour under formal minority governments and minimal winning governments were not statistically significant, providing further evidence for the argument that formal minority governments are ‘majority governments in disguise’.

The role of policy distances to the government in opposition parties’ decision to support or oppose the government was also expected to increase under a substantive minority government. The results of the moderation analyses, however, show that the effect of policy distance only differs significantly between substantive minority governments and majority governments in Canada (Model 2). Here, policy distance only has a negative and significant effect under substantive minority governments in Canada but not under majority governments (see Appendix C, Table C2). In the other country models and the pooled model, however, there was no statistically significant moderation effect of cabinet type.

The trade-offs between the opposition objectives were expected to tilt further into vote-seeking behaviour when opposition parties feel more observed by voters. Under those circumstances, opposition parties’ voting behaviour should be informed more strongly by their policy positions, because the signal that their position in parliament sends to voters is more effective. When a party is positioned further away from the government, voting against the government comes with greater electoral benefits as it enables opposition parties to distinguish themselves from the government on their core issues. When positioned close to the government it is harder to oppose credibly and easier to justify support. The empirical analyses provide no support for these expectations (Model 3). Neither in the pooled model nor in the country models was the moderation effect of issue saliency and policy distance statistically significant. The main effect of issue saliency differs between countries. Only in the Dutch Tweede Kamer does issue saliency lead to a lower probability of voting in support of government bills (see Appendix C, Table C1). Contrarily, in Denmark and the United Kingdom, high issue saliency was associated with higher levels of support for government legislation.

Opposition parties should care not only about winning the next elections and exerting policy influence but also about replacing or joining the incumbent government. Increasing internal tensions within coalition governments could be a fruitful way to stimulate an early government alternation. Therefore, opposition parties were expected to oppose government bills on policy areas on which coalition parties were divided, in order to highlight the painful compromises coalition parties make and embarrass them. The analyses reveal support for this hypothesis only in the Dutch case. Here, for each standard deviation increase in the policy range within the government, the probability of voting in favour of government bills decreases by 1.4 percentage points. In Denmark, however, the coalition policy range did not have a significant effect (see Appendix C, Table C1).

The strategy of driving a wedge between coalition parties was hypothesized to be used to a different degree depending on the specific institutional context and the position of opposition parties within that context. First of all, office-seeking behaviour should be more attractive for those opposition parties that are likely to enter future governments than for parties that are or seem to be in opposition permanently. For those latter parties, policy and vote objectives should usually prevail. Second, the patterns of government alternation within a political system should determine the strategy employed by opposition parties with office aspirations. In systems of wholesale alternation, such as Canada, Denmark and the United Kingdom, attacking the incumbent government is the best way to stimulate government instability and a sooner replacement of the government. In systems of partial alternation, opposition parties should moderate their attitude towards the government to signal trust and responsibility towards future coalition partners who are currently in government. The analyses reveal that the more credible the position of the alternative government is for an opposition party, operationalized as the degree of government experience, the more the opposition parties will vote against government bills. This effect is only substantial and statistically significant in the systems of wholesale alternation and not in the country of partial alternation (the Netherlands).

Furthermore, moderation analyses of the Danish and Dutch cases show that Danish opposition parties aspiring for office use the wedge-driving strategy more, whereas their Dutch counterparts use the wedge-driving strategy less (see Table 4). This supports the expectation that the context of government alternation patterns affects the behaviour of opposition parties who aim to enter the government. In Denmark, the more government experience opposition parties have, the stronger the negative effect of coalition issue divisiveness is on support for government bills. In the Netherlands, the more government experience opposition parties have, the weaker the negative effect of coalition issue divisiveness is on support for government bills.

Table 4. Multilevel Logistic Regression Analyses of Opposition Support for Government Bills (With Government Experience and Government Policy Range Interaction)

Notes: The reference category for cabinet type is minimal winning.

*** p < 0.001, **p < 0.01, *p < 0.05, ^†p < 0.1.