Introduction
Governments in parliamentary democracies can come in various shapes and formats. As government formation is a post-electoral bargaining process among political parties, voters do not have a direct say which political parties ultimately form coalitions, divide up ministerial portfolios, support government policy, or remain in opposition. A specific outcome of post-electoral bargaining or government terminations are minority governments, i.e., instances in which cabinet parties do not muster a parliamentary majority. Such governments may present a challenge for accountability: the parties that hold power in cabinet and ministerial portfolios are not identical to those that pass legislation in parliament, since additional legislative support is – by definition – required. While they have been historically always prominent in parliamentary democracies, with around one in three cabinets being a minority government (Field and Martin Reference Field and Martin2022), this government type provides flexibility in post-election bargaining when parliaments are polarized. It allows for the formation of governments that may not be feasible politically if some of the legislative support parties were formally included as cabinet parties.
Our pre-registered study examines citizens’ knowledge and perceptions of such minority governments. Studies have repeatedly shown that minority governments can work effectively in parliament (e.g., Field and Martin Reference Field and Martin2022; Strom Reference Strom1990), yet we lack a more fundamental understanding of their representational implications at the voter level. Specifically, we ask two questions: (1) whether citizens know what a minority government is and (2) how this knowledge affects their support for and perceptions of minority governments. The direction of the relationship between knowledge and support is an empirical question: on the one hand, knowledge about minority governments and the legislative difficulties they face may decrease support for them. On the other hand, knowledge about how minority governments can still effectively govern and involve compromise may increase support.
We study these questions in Denmark, Germany, and Sweden and find that citizens have limited knowledge about minority government in a parliamentary system. In all countries, around 20% of citizens can correctly identify a minority government, levels that are comparable to the knowledge of the political role of the parties in the country. Despite the differences in historical prevalence of minority governments, citizens in all countries appear to prefer majority generally over minority governments. Those knowledgeable about minority governments, in fact, express a stronger preference for majority governments, while those not knowledgeable do not prefer minority governments more, but simply are unsure which government type they prefer.
Despite the general preference for majority governments, our study shows that voters associate minority governments more with the impression that such governments listen more to the demands of all voters. While majority governments are seen as more stable, reliable, and able to solve societal conflicts, minority governments are also seen to cooperate well with the opposition. These effects are even stronger for those who can correctly define minority governments. Voters appear to associate distinct positive and negative representational aspects with minority governments, thus possibly enabling politicians to emphasize such traits when defending or criticizing these government types in parliamentary democracies.
Voter perceptions of minority governments
The formation of minority governments is the result of a particular legislative bargaining outcome with two components (Kalandrakis Reference Kalandrakis2015; Strom Reference Strom1990). First, one or more parliamentary parties with strong policy-seeking motivations are able to influence legislation by offering their parliamentary votes to a government that does not control a legislative majority on its own. Such arrangements may be more or less formalized (Krauss and Thürk Reference Krauss and Thürk2022; Thürk and Krauss Reference Thürk and Krauss2024). Second, even though more legislators than those of the government parties are required to pass legislation, only the governing parties control ministerial offices. Ultimately, while it is established that minority cabinets last shorter than cabinets with legislative majorities, they tend to do equally well with regard to implementing policies and fulfilling electoral pledges (Artés and Bustos Reference Artés and Bustos2008; Klüver and Zubek Reference Klüver and Zubek2018; Moury and Fernandes Reference Moury and Fernandes2018).
How are such governing arrangements perceived by voters? The early literature on clarity of responsibility (Powell Reference Powell2000; Powell and Whitten Reference Powell and Whitten1993) suggested a negative implication, as accountability would be one of the lowest for situations with minority governments. Power is shared, legislative coalition-building becomes necessary to get things done in parliament, and voters may therefore be less able to attribute responsibility correctly to government, support, and opposition parties. The empirical evidence, however, suggests that the attribution of responsibility appears to work similarly in majority and minority contexts (Dassonneville and Lewis-Beck Reference Dassonneville and Lewis-Beck2017; Fisher and Hobolt Reference Fisher and Hobolt2010; Hobolt, Tilley and Banducci Reference Hobolt, Tilley and Banducci2013; Vowles Reference Vowles2010). In the Canadian context, Dufresne and Nevitte (Reference Dufresne and Nevitte2014) have shown that citizens can have an explicit preference for minority governments, at least when such governments favour their preferred party. In a survey experiment in Germany, Matthieß and Stecker (Reference Matthieß and Stecker2025) similarly show that while citizens prefer majority governments over minority governments, the potential for policy influence of one’s preferred party increases the support for minority governments.
In the context of election studies, survey researchers have asked respondents explicitly about opinions towards minority governments, in particular when such governments were a possibility in the post-election government formation process. Dufresne and Nevitte (Reference Dufresne and Nevitte2014) finds that 40–50% of respondents in Canada favour minority governments, while 20–25% are against such arrangements. In Germany, the Politbarometer asked respondents during the lengthy government formation period after the 2017 federal election, whether they preferred a CDU/CSU minority government or new elections in case the grand coalition would not be possible (eventually, a grand coalition was formed). Respondents were split equally between these options, but remarkably, less than half of CDU/CSU voters expressed a preference for a CDU/CSU-led minority government – giving all ministerial offices to the respondents’ preferred parties – over new elections (Forschungsgruppe Wahlen, 2019). In contrast, the minority government possibility had the greatest support among opposition party voters, particularly among voters of the Left, Greens, and Social Democrats. Other studies on the national level have asked about support for minority governments before elections, with generally strong opposition among the electorate (e.g., Berger et al., Reference Berger, Jung, Roth, Gibowski, Kaase, Klingemann, Küchler, Pappi and Semetko1995; Berger, Jung and Roth, Reference Berger, Jung and Roth2003; Jung, Laemmerhold and Wolf, Reference Jung, Laemmerhold and Wolf2014). On the sub-national level in Germany, support for minority governments is highly context dependent, as respondents react to previous experiences and the potential parties involved in such governments (e.g., Forschungsgruppe Wahlen, 2014; Forschungsgruppe Wahlen, 2012). Overall, support for minority governments seems to depend on contextual factors, such as the perception of other options, the opportunity for policy-influence of one’s own party, the fear of failed government negotiations, and previous exposure to such governments.
Some authors have explored the perceptions of government types in more detail. Fortunato et al. (Reference Fortunato, Lin, Stevenson and Tromborg2021) find in a survey experiment that respondents attribute higher policy influence to parties that are designated as “support parties” compared to opposition parties. In other words, when priming voters to think about the existence of support parties, then voters also believe such parties are more influential. However, we do not know whether voters respond more to a priming effect of a “support” party or whether citizens are actually knowledgeable about the fact that the cabinet is without its own majority in parliament.
Political knowledge is widely considered an essential part of a functioning democracy. Political knowledge can refer to facts about the political system or policies, and moreover relate to general static facts or facts that relate to recent political developments (Barabas et al. Reference Barabas, Jerit, Pollock and Rainey2014). Knowledge about government composition refers to a dynamic component (one has to remember the last election results) and to general aspects (which parties have formed the government). Studies have shown that citizens with greater political knowledge in this domain are better able to understand coalition dynamics and hold parties accountable (Fortunato and Stevenson Reference Fortunato and Stevenson2013). Without knowledge about government composition, for instance, voters struggle to reward or sanction parties appropriately at elections. Not only is there a relatively large share of voters who do not know the basic allocation of power in parliamentary systems among parties (around one in three, on average), but the estimate of economic voting increases substantially once such voters are excluded from the analysis. Other studies have furthermore shown that past experience with specific governments increases political knowledge. Fortunato, Stevenson and Vonnahme (Reference Fortunato, Stevenson and Vonnahme2016) show that when left-right positions predict the composition of coalitions, voters also have a greater level of partisan left-right knowledge (in terms of placing parties relative to each other). Similarly, we would expect that past experience with minority governments also increases knowledge about the situation as well as perceptions.
Our study builds upon this work and investigates general static knowledge about the parliamentary system, specifically about minority governments, as a feasible outcome of coalition bargaining. We examine the determinants of this knowledge, focusing on individual-level factors (i.e., socio-demographic characteristics), party-level factors (i.e., whether respondents are government supporters), and system-level factors (i.e., if they have experienced a minority government as voters).
Data and methods
We investigate preferences for and perceptions of minority governments, and the impact of political knowledge, on the basis of a survey in Denmark, Germany, and Sweden, with
$N = 1,000$
respondents each, fielded in January 2023 with Lucid (Proksch and Wäckerle Reference Proksch and Wäckerle2025). The three cases represent countries with different exposure to minority governments at the national level. While Germany has never experienced a minority government in a post-electoral context at the national level, there is frequent discussion about the potential formation of minority government following federal elections and frequent formation of such government types at the regional level. Denmark and Sweden frequently experience minority government formation. At the time of the survey, all three countries were governed by a coalition of parties. In Germany, a coalition of Social Democrats, Greens, and Liberals was formed following the 2021 election, with the parliamentary opposition being divided into parties on the centre right (CDU/CSU), far right (AfD), and far left (Linke). In Denmark, a coalition by the Social Democrats, Liberals, and Moderates took office in 2022, bridging the traditional divide of left-right bloc politics. The government was supported by the Social Liberals as well as members from the Faroe Islands and Greenland. In Sweden, a minority coalition government formed between Moderates, Christian Democrats, and Liberals following the 2022 election. The far-right Sweden Democrats, for the first time, constituted the official support party of the government, leaving the opposition in parliament on the left of the political spectrum. This case selection allows us to explore whether Danish and Swedish citizens are more knowledgeable about minority governments compared to Germans, and if such knowledge also is associated with different opinions about the governments.
In the survey, we first ask for demographic information, political ideology, and vote choice. Subsequently, the respondents are asked to select correct statements about minority governments from a list to test the level of knowledge about this form of government. We then randomly show half the respondents the correct definition, while the other half get no information treatment. We ask two outcome questions: first, respondents can state a direct preference between minority and majority governments in a case in which both are theoretically possible. Second, we ask whether certain government traits are more connected to majority or minority governments. The full questionnaire is in the online Appendix H.
The survey employed quotas for age groups, gender, region, and education. We excluded speeders, who finished the survey in under three minutes, from survey completion. The median answer length was about six minutes. While there were no quotas on voting behaviour at the previous national election, self-reported recall of vote choice largely corresponds to the respective election results (details are in online Appendix B). The survey was preregistered with osf.ioFootnote 1 and received ethics clearance from the ethics board of the Faculty of Management, Economics and Social Sciences at the University of Cologne.
Results
Knowledge of minority governments
We first investigate which respondents in each country are knowledgeable about minority governments. We presented respondents with six statements and asked them to select the ones that they associate with minority governments. We also asked them to answer “don’t know” if they had never heard of the term “minority government”. Two of the six statements contained the correct definition: government parties do not control the majority of seats in parliament and the government relies on support from some members of the opposition in parliament to pass legislation. The other statements indicated that government parties received less than half the votes in the election, that the government includes parties that represent minority groups in society, that the government does not include the largest party in parliament, or the option to select none of the statements.
The answer option “less than half the votes” is not a sufficient condition for defining a minority government, as majority governments could also have been elected with less than majority support of the electorate. In fact, Figure C1 in Appendix C shows the prevalence of such governments (seat majority with vote share minority) since 1990 using data from Parlgov (Döring and Manow 2024). In many countries (for example, majoritarian systems such as Australia, France, or the UK), this government constellation makes up the majority of all governments. In only two countries (Denmark and Austria) this has not happened since 1990, in Austria due to their proportional system and in Denmark, as they have not had majority governments since 1990. Thus, this answer option separates out respondents who are aware that minority governments are defined solely in the legislative context and those who do not understand the technical difference between a minority of seats and a minority of votes.
Figure 1 shows the estimated share of respondents that can correctly define a minority government. We code respondents as knowledgeable if they give at least one correct answer (seat share or cooperation) without an incorrect answer (minority groups + largest party). In the first line (wider definition), we additionally count the option naming a minority of votes as correct, while we disregard respondents’ selection of the vote share option in the second line (narrow definition). In other words, respondents need to select at least one correct definition without selecting an incorrect one in order to be coded as knowledgeable. Respondents that answer “Don’t know” are coded as not knowledgeable about minority governments.Footnote 2 The results show that around 20% of respondents in all three countries know what a minority government is, with surprisingly few differences between the countries despite different exposures of the electorates to minority governments.Footnote 3 This estimate of the true knowledge is calculated by putting the share of correct answers in relation to the share of respondents that could have guessed the answer. Assuming the same distribution of numbers of answers picked as in the actual survey, about 31% of respondents would have correctly guessed the definition (41% for the wide definition). The figure furthermore compares these knowledge levels to two other common measures for political knowledge (Fortunato and Stevenson Reference Fortunato and Stevenson2013), derived from our same survey. In all three countries, 65–80% of respondents can correctly identify the political party of the prime minister, while 45–60% can name all government parties (without support parties) and 15–35% correctly identify all parties’ roles as government or opposition parties. Support parties are ignored for this calculation because we did not give respondents a separate prompt for support parties, as we did not want to indirectly provide information about minority governments. In sum, political knowledge about minority governments is comparable to that of correctly identifying the role of all parties in a country. We interpret these findings as reassuring: around one quarter of all respondents have a correct understanding about a common post-electoral outcome that is solely due to inter-party bargaining. Meanwhile, the substantial number of respondents correctly identifying PM, government, and parliamentary parties represent very high knowledge of political facts among respondents.
Figure 1. Political knowledge about minority governments and government composition.
In the Appendix, we present the different answers in a disaggregated form (Figure C3). Respondents chose most frequently the option that the government needs to rely on support from the opposition. However, a similar share of respondents selected the vote share and seat share option, indicating indeed a considerable degree of perceptual confusion between these two options. Only very few respondents selected the wrong response options, although the option that a minority government does not include the largest party in parliament was selected by almost 25% of respondents in Sweden, where at the time of the survey the Social Democrats and Sweden Democrats constituted the largest and second-largest party respectively, but are not part of the minority government.
Finally, we investigate the individual determinants of knowledge about minority governments. As Figure 1 indicates, the average share of people who can correctly define a minority government across the three countries is very similar. In order to investigate the predictors of knowledge in more detail, we run a logistic regression with individual-level covariates to explain variation in knowledge, with the dependent variable being whether a respondent correctly defined a minority government or not (narrow definition). As predictors, we include individual-level covariates including education, gender, age, political interest, ideology, and general political knowledge. We also add variables indicating whether the party a respondent voted for had prior minority government experience and is currently in government or the opposition.
The results of this regression are in Table C1 in Appendix C. Model 1 (without general political knowledge) shows that respondents with lower general education, women, and those with little political interest are less likely to correctly define minority governments. Meanwhile, older respondents are more likely to do so. However, this effect might be partly due to some respondents, for example, women (Ferrín, Fraile and García-Albacete Reference Dassonneville and Lewis-Beck2017), being more likely to select the “Don’t know” option, partly out of an aversion to answering the question if they have doubts about their answer (for more information, see Appendix C.5). In model 2, we add a variable for general political knowledge. We measure general political knowledge by assessing whether a respondent correctly identified the prime minister’s party, as this is fairly comparable across countries (the task of correctly naming all parties in parliament and their roles is much harder in Denmark than in Germany as a result of a higher number of parties in parliament). This variable is strongly related to knowledge of minority governments: the predicted probability of correctly defining minority governments is 0.48 for those with knowledge of the prime minister party, while it is 0.29 for those who do not know the party.
Once we control for general political knowledge in this way, some socio-demographic effects become weaker: the estimates for age are considerably smaller, indicating that some of this effect is due to general political knowledge being higher among older voters. However, the other effects persist, which indicates that these variables are connected to knowledge about minority governments in a way that goes beyond general political knowledge.Footnote 4
In model 3, we add the vote choice variables: while non-voters are slightly less likely to know what a minority government is, there are no statistically significant differences between government, opposition, and support party voters. Additionally, prior minority government experience is not related to higher knowledge of minority governments. Once we control for all these variables, differences between the countries remain small and not statistically significant beyond the 90% level. A similarly educated and knowledgeable person in all three countries is expected to know what a minority government is at roughly the same rate.
Finally, in Appendix G , we test whether providing information about minority governments to respondents changes their subsequent answering pattern and find that this information treatment does not work.Footnote 5 For survey research, this points towards an overall limitation: many respondents are unaware of complex political phenomena such as minority governments and simple disclaimers in surveys are unlikely to inform them sufficiently. Researchers should therefore measure knowledge of respondents and analyse effects within knowledge subgroups.
Support for minority governments
After establishing knowledge differences in the electorate on minority governments, we investigate the consequences for support for and general perceptions of minority governments. We asked respondents the following question:
Please think about a hypothetical situation in which a government is formed. Please ignore for a moment the actual parties that would be members of that hypothetical government. In a situation in which both a majority and a minority government would be feasible, which of the two would you prefer?
Overall, Figure 2 shows that respondents in all countries prefer majority to minority governments, with only 13% of all respondents preferring minority governments, 60% preferring majority governments, and 27% being undecided. The results also show that the general preference for majority governments is highest in Sweden with over 70% of respondents, while Denmark has the highest support for minority governments with 20% preferring such cabinets. German respondents are in between Danish and Swedish respondents in terms of their preference for minority government. In short, respondents express general support for majority governments independent of whether minority governments are the norm (Denmark), common (Sweden), or rare (Germany) in their respective countries. However, there are still notable differences between the countries, with the strongest support for minority government in Denmark, followed by Germany and Sweden.Footnote 6
Figure 2. Preferences for minority government across countries.
In Table E1 in Appendix E, we present regression results for preferences for minority government. The first model only includes socio-demographic variables, while the second model adds general knowledge and the third additionally party-level factors. The results show that Danes, young voters (under 45 years old compared to those over 55 years old), men, respondents with a high interest in politics and opposition party voters are more likely to prefer minority government. Prior minority government experience of a respondent’s party has no effect on preferences.Footnote 7 However, Table E2 shows the same regression with preference for majority government as the dependent variable and here we find that voters of parties previously in minority government slightly prefer majority government.
Knowledge about and preference for minority government are related: Tables E1 and E2 show that voters without knowledge of minority governments prefer minority government over majority government more than those that can correctly define them. The predicted values from model 3 in both tables are shown in Figure 3. Support for minority governments is less than 12% for those with knowledge of minority governments, while this increases to almost 15% for those that do not know what minority governments are. Knowledge about minority governments thus appears not to increase – but rather decrease – support for them.
Figure 3. Preferences for minority government by knowledge.
Perceptions of minority governments
Our final investigation examines voters’ perceived positive and negative associations of minority governments relative to majority governments. We present respondents with a list of statements and ask them about their opinion of which of them apply more to minority governments, majority governments, or equally to both. We select statements that relate to the efficiency and representational dimensions of governments. Finally, we distinguish between respondents who are knowledgeable about minority governments and those that are not. Figure 4 presents average results across all three countries, with the dots denoting the average respondent, triangles the knowledgeable respondents, and squares those without knowledge. We furthermore group the nine statements based on a factor analysis of the responses, with correlated responses appearing in the same group (see details in Appendix F).
Figure 4. Opinions about minority and majority governments.
First, respondents overall hold several negative opinions about minority governments. Not only do respondents believe that such governments are less able to solve intense societal conflicts, but they consider them also to perform worse, with more instability and less legitimacy of government decisions. In a similar vein, respondents associate the statement that democracy is strengthened more with majority governments than with minority governments. Voters, moreover, think that majority governments are a more reliable partner at the European and international levels compared to minority governments, despite evidence of both government types being able to enter international commitments in similar ways (Oktay Reference Oktay2022). This suggests that minority governments are more under scrutiny by voters when negotiating internationally, despite being similarly effective. A correlated statement with the international commitments is that respondents indicate that majority governments are less likely to change their stances in line with public opinion. While the statement can be interpreted as an expression of responsiveness in a representational aspect, respondents appear to interpret this in a negative way, as it is correlated with less reliability at the international level. Thus, holding course against public opinion is associated with majority cabinets, while minority governments are considered more volatile.
In contrast to these negative perceptions, on some dimensions minority governments fare better with voters. Voters acknowledge that the government cooperates better with the opposition than majority cabinets, confirming the positive effects of perceived opposition influence and support for minority governments. Importantly, minority governments are deemed to listen more to the demands of all voters compared to majority governments. This suggests that voters do think that compromise-making in parliament and outside of cabinet may lead to better representational outcomes.
Finally, our results reveal that respondents with knowledge about minority governments have consistently stronger opinions about them, both in the positive and negative directions. For example, they associate instability much more with minority governments than those who cannot correctly define them, but also think that such governments cooperate better and listen to demands of all voters. The only items for which there are no differences are the legitimacy, corruption, and democracy claims. These are the items for which respondents see the smallest difference between majority and minority governments.
In short, voters see minority governments as responsive and cooperative, but unstable and less reliable in international situations. When disaggregating the results by country, we find that while the overall effects are comparable, the perceptions of minority governments are more positive in Denmark than in Germany and Sweden.
Conclusion
Our conclusions about knowledge of minority governments are mixed. First, only one in five voters can correctly define a minority government. This knowledge is comparable to that of correctly identifying the role of all political parties in the respective country. Second, more knowledge does not translate into higher support for minority governments. In all countries, independent of the prevalence of minority governments, voters prefer majority governments to minority governments, especially in Sweden. Knowledge about minority governments, furthermore, seems to be associated with an even stronger preference for majority cabinets. Third, with regard to perceptions, voters see minority governments to have better process in the policy-making domain – cooperating better with the opposition and listening more to the demands of all voters –, but worse output and reliability. Notably, voters who know what minority governments are have stronger perceptions about them, both in the positive and negative directions.
From an elite communication perspective, this yields interesting implications in how such arrangements should be communicated by parties in minority governments to supporters and voters more generally, but also for opposition parties who can point out specific critiques that are likely to resonate well with voters. However, we note that our study only relies on three countries, raising the question about the impact of country-level factors on support and perceptions. Future research should therefore study these question in a broader sample of countries.
Our findings also have important implications for survey research: Knowledge among respondents, especially on complex issues, should not be assumed, but rather needs to be tested and taken into account when analysing treatment effects. For instance, the perceived legitimacy of decisions by minority versus majority governments may depend on whether respondents understand what a minority government is. This is compounded by the effect that it does not appear to be sufficient to simply tell respondents what key concepts mean and hope to even out the knowledge level among respondents within the survey. Therefore, it is important to know the level of understanding among different groups of respondents to make inferences about heterogeneous treatment effects by prior knowledge. Moreover, knowledge is not randomly distributed in the public. Without taking respondents’ understanding of key concepts into account, we could draw wrong conclusions from survey results. For example, if one group prefers minority to majority governments and another has the opposite preference, we would not know whether these are differing preferences or differing understandings of minority governments. We leave the answers to these questions for future research.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S147567652510025X
Data availability statement
The dataset is available at the GESIS data repository (https://doi.org/10.7802/2917). The code for replicating the study is available under Supporting Information.
Acknowledgements
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 101002115). We would like to thank Lukas Hetzer, Paula Hoffmeyer-Zlotnik, Hauke Licht, and Jasmin Rath for feedback when designing this survey. We are grateful to the participants and discussants for their valuable feedback at the Political Representation and Behavior Workshop in Cologne in 2023 and at the EPSA 2023 and APSA 2023 annual conferences. We thank John Högström for comments on an earlier version of the paper, Frederik Hjorth for support with the Danish translations of the questionnaire and Mia Hoymann for excellent research assistance.
Open access
