When Does Party Cooperation Inform Voters Judgments of Policy Proximity? A Cross-National Analysis of Heuristic Use in Multi-Party Democracies

Inter-Party Cooperation as a Policy Heuristic

In multiparty systems, relationships between political parties are shaped not only by electoral competition but also by patterns of cooperation. A substantial body of research shows that voters perceive parties’ left–right positions to be closer together when those parties join formal governing coalitions (Adams et al., 2016; Fernandez-Vazquez, 2014; Fortunato & Adams, 2015; Fortunato & Stevenson, 2013; Spoon & Klüver, 2017). This effect grows stronger the longer coalitions last (Fortunato, 2012) and also appears when cooperation is informal, such as through legislative agreements or coordinated policy action (Adams, Weschle, & Wlezien, 2021; Falcó-Gimeno & Muñoz, 2017; Santoso, 2025a).

These patterns are consistent with the idea that voters use changes in partisan cooperation as a cue for policy change, perceiving parties that work closely together as holding more closely aligned preferences. While most of this work conceptualizes proximity in broad left–right terms, the same logic can apply at the level of specific policies, where signals of cooperation may be more immediately relevant to voters’ views on concrete issues such as immigration, redistribution, or climate change.

If partisans believe their party is moving closer to an out-party, partisan rationalization theory predicts that they will adjust their own views accordingly. Because many citizens take policy cues from their preferred party (Lenz, 2012; Zaller, 1992) and view that party as a core component of their social identity (Green et al., 2002; Huddy et al., 2015), they are motivated to remain aligned with it. When cooperation signals that the two parties are closer in policy terms, partisans may shift their own stance toward the out-party’s position to preserve loyalty and cognitive consistency, avoiding the discomfort of holding views at odds with their party (Goren et al., 2009; Taber & Lodge, 2006).

A recent study illustrates this mechanism. Santoso (2025b) finds that when identifiers of non-radical right parties (non-RRPs) perceive their party cooperating with a radical right party (RRP), they move their own positions, particularly on highly salient issues such as immigration and cultural diversity, closer to the RRP’s stance. Crucially, this occurs even without specific objective policy agreements between their supported party and the RRP, indicating that the cooperation itself is enough to trigger partisan rationalization.

The same logic, however, can be extended across the political spectrum. When any two parties cooperate, voters may interpret this as evidence of shared policy ground and, if they are partisans of one of the parties, adjust their own policy positions toward those of the out-party. In this way, inter-party cooperation functions as a policy heuristic: an observable, behavior-based cue that not only signals policy alignment but also activates identity-based and motivational processes that drive partisan rationalization across diverse political contexts and policy domains. By heuristic, I mean a rule that maps observable informational inputs (cues) and the weights placed on them onto a target inference or belief with less effort than more deliberative alternatives (Gigerenzer et al., 1999; Lau & Redlawsk, 2006; Shah & Oppenheimer, 2008). Here, perceived cooperation between parties is the cue and the voter’s inference about policy proximity to the out-party is the target.

Importantly, this mechanism does not require actual policy convergence. Since inter-party interactions are frequently documented by the media (Adams, Bernardi, & Phillips, 2021, Adams, Weschle, & Wlezien, 2021; Santoso et al., 2024), voters can observe and assess cooperation between specific party pairs. This visibility allows them to infer an out-party’s likely policy positions even in the absence of objective change. As a result, cooperation can shape mass opinion even when it reflects symbolic gestures rather than objective policy agreements. ⁴

Based on the explanation above, I hypothesize that:

H1

Greater perceived cooperation between a partisan’s preferred party and an out-party will be associated with a smaller gap between the partisan’s own policy position and the out-party’s position on that issue.

While the expectation in H1 follows directly from the partisan rationalization mechanism, such cooperation will not always serve as an equally strong or reliable signal for reducing the perceived policy distance between the partisan’s own position and that of the other party. In some contexts, partisans may dismiss cooperation as a matter of convenience or parliamentary necessity. In others, they may view it as compelling evidence of shared policy ground. The case of cooperation with radical right parties discussed above illustrates this point: when the other party “owns” a salient issue such as nativism, even perceptions of cooperation in general can be politically meaningful and more likely to activate the partisan rationalization process.

This raises a broader question: when is inter-party cooperation most likely to operate as an effective heuristic for judging policy proximity? Partisan rationalization explains why cooperation works as a heuristic in the first place: to preserve consistency with their in-party, supporters interpret cooperation as evidence that the out-party must be closer to their own position. Yet rationalization alone does not tell us when this heuristic will be influential. Previous works on heuristics use suggest that the usefulness of a cue depends on the informational environment in which voters operate. Using an ecological rationality framework, Fortunato and Stevenson (2019) show that voters rely on heuristics to infer legislators’ behavior when partisan cues are sufficiently informative. Similarly, Fortunato et al. (2021) demonstrate that citizens attribute policy-making influence on parties using observable signals such as their roles in government and party size. These studies highlight that voters often rely on readily available political cues when direct information about party positions and policy influence is difficult to obtain. Using the same ecological rationality framework, I examine when patterns of inter-party cooperation function as cues for inferring policy proximity. While these works focus on cues derived from party labels or government roles, this paper considers a different type of cue — perceived interactions between one’s preferred party and an out-party—and when voters use these interactions to infer policy proximity across multiple issues.

Ecological Rationality: Conditions for Using Partisan Cooperation as a Policy Heuristic

As Gigerenzer (2008) and Gigerenzer and Gaissmaier (2011) argue, heuristics should be evaluated not by how closely they approximate full-information reasoning, but by how well they perform in the specific environments where decisions are made. In the context of inter-party cooperation, this perspective highlights three conditions that determine whether voters will use the cue: cheapness (the cue is easy to access or recall), simplicity (the cue is straightforward to interpret), and accuracy (the cue is informative about the proximity of out-party’s likely policy position). ⁵

First, a heuristic is “cheap” if the values of relevant cues are widely known or easy to obtain. In the case of inter-party cooperation (i.e., collaboration between an individual’s preferred party and an out-party), this means that voters can easily recall or observe patterns of collaboration between parties. Modern media environments, particularly in multiparty democracies, make this possible. Mainstream outlets routinely report on coalition negotiations, co-sponsored legislation, joint press conferences, and other visible signals of collaboration (Adams, Weschle, & Wlezien, 2021; Santoso et al., 2024). The rise of partisan media has amplified this accessibility: outlets sympathetic to a voter’s preferred party often frame cooperation as evidence of pragmatic problem-solving or strategic acumen, while hostile outlets may portray it as betrayal or ideological compromise (Jamieson & Cappella, 2008; Levendusky, 2013). Both frames make the cooperation salient, and repeated exposure embeds it more firmly in voters’ memories, lowering the cost of retrieving this information when forming judgments about policy.

The ease to which voters can access and recall information about cooperation, however, is not uniform. While contemporary media ecosystems, especially partisan outlets, can make instances of inter-party cooperation more visible, individuals still vary in their ability and motivation to take in and store this information. Politically knowledgeable individuals are more likely to follow news about parliamentary developments, recognize patterns of inter-party collaboration, and retain this information over time (Prior, 2018). Their interest in politics means that cues about cooperation slot easily into an existing mental schema, making retrieval almost effortless. In contrast, less knowledgeable voters may remain unaware of many cooperative episodes unless they are unusually prominent or repeatedly emphasized in the media. For them, the cue is less “cheap” because it must be noticed afresh each time rather than drawn from a well-established store of political knowledge. ⁶

Under these conditions, the cooperation cue is not only cheap but also more likely to be used in making policy proximity judgments. For politically knowledgeable individuals, the mental cost of recalling instances of cooperation is minimal, meaning the cue can be accessed quickly and deployed in low-effort reasoning about whether an out-party’s policies are close to their own. Accordingly, the association between perceived general cooperation and perceived policy proximity should be stronger among politically knowledgeable individuals, for whom the cooperation cue is especially easy to access and recall.

Second, a heuristic is simple if the mapping that applies the weight of a cue to generate an inference is cognitively relatively undemanding for the average person. ⁷ While cheapness is about the cost of accessing or recalling the cue, simplicity refers to the cost of processing it once it is in mind. In ecological rationality terms, simplicity refers to the minimal cognitive processing required to translate an observed cue into a belief or judgment (Shah & Oppenheimer, 2008). The simpler the cue–inference mapping, the less mental effort is required, and the more likely the heuristic will be used in practice (Gigerenzer & Brighton, 2009).

In the case of partisan cooperation, simplicity increases when voters can interpret the meaning of cooperation quickly and with reference to an existing, well-understood political framework. One such framework, which is deeply embedded in mass political cognition, is the left–right ideological spectrum. Across many democracies, the left–right scale functions as a “super-issue”, serving as a cognitive shortcut that summarizes a party’s general ideological orientation and, by extension, its likely position on a range of policy issues (Huber & Inglehart, 1995; Knutsen, 1997; Mair, 2009). Voters regularly use left–right placement as a shortcut to organize their own beliefs, assess other parties, and predict where those parties stand without consulting detailed policy platforms.

From this perspective, the ideological extremity of the out-party makes the cooperation cue simpler to interpret. When the out-party occupies a far-left or far-right position, its ideological identity is distinctive and highly diagnostic: voters already have a clear sense of what such a party stands for across many issue domains (Adams et al., 2006; Wagner & Meyer, 2017). ⁸ Observing cooperation with such a party therefore makes it straightforward to infer some level of policy proximity, since such cooperation is unlikely to be dismissed as mere procedural necessity or symbolic bipartisanship. The out-party’s extremity thus serves as an anchor — its positions are already sharply defined, making any visible cooperation stand out as a meaningful signal of common ground that voters can act on without needing to process detailed legislative records or issue-specific compromises.

In contrast, cooperation with a centrist or moderately positioned out-party is more ambiguous. Centrists often share partial policy overlaps with parties on both sides of the ideological divide, so their cooperation does not clearly signal genuine policy alignment. They also tend to have weaker valence attributes — lower perceived competence, integrity, or cohesion — which further dilutes the informational value of their cooperation (Zur, 2021a, 2021b). Voters may therefore interpret such cooperation as strategic flexibility or routine compromise rather than substantive agreement, making the cue noisier and less useful as a heuristic.

Thus, the ideological extremity of the out-party reduces the complexity of the inference process. Based on the discussion above, I expect cooperation with ideologically extreme out-parties to have a stronger effect on perceived policy closeness than cooperation with centrist ones.

Finally, a heuristic is considered accurate when the cue it relies on co-varies reliably with the outcome it is meant to signal. In the ecological rationality framework, individuals rely on cues that, over time, have proven statistically informative in their environment (Gigerenzer et al., 1999; Pohl et al., 2017). For example, a voter may use perceived inter-party cooperation as a shortcut to infer policy proximity between their preferred party and an out-party. ⁹ Put simply, a cue is accurate when it consistently co-varies with the outcome of interest. ¹⁰

Importantly, this does not imply issue-specific cooperation. In the empirical design that follows, I measure cooperation in general, while policy proximity is assessed across a wide range of issue domains. The heuristic is judged accurate to the extent that general perceptions of cooperation predict convergence on policy scales. This allows for meaningful variation: in some domains, cooperation cues may reliably indicate proximity; in others, they may not.

A central moderator of cue accuracy is issue-level polarization. On polarized issues, party elites tend to adopt internally coherent positions that sharply diverge from those of competing parties (Gidron et al., 2020; Levendusky, 2010; Mehlhaff, 2025). In such environments, any signal of cross-party agreement carries substantial political risks as partisans may react with backlash (Golder, 2006), and party leaders risk reputational harm among ideological allies (Martin & Vanberg, 2011). Because symbolic or merely procedural cooperation is politically costly, it is less likely to occur unless it reflects substantive alignment. As a result, when voters perceive cooperation and evaluate elite positions on polarized issues, those perceptions are more likely to reflect real convergence. In these cases, the cooperation cue is more diagnostic (i.e., more accurate).

By contrast, on issues with low polarization, elite positions are less distinct, and the political costs of cooperation are lower. Parties frequently coordinate for reasons unrelated to ideological alignment: passing routine legislation, resolving technical disputes, or engaging in strategic signaling. Voters may still perceive cooperation, but those perceptions are less informative of actual proximity. The heuristic, in these contexts, becomes noisier and less reliable.

This logic is reinforced by recent models of coalition governance. For example, De Marchi and Laver (2023) show that parties often construct governing agendas by tabling contentious issues and negotiating only in domains where agreement is feasible. Cooperation, in this view, does not imply alignment on all issues, instead, it reflects strategic issue selection under constraints. When voters perceive general cooperation, they may not know which issues are involved, but the underlying production of cooperation is not random. In polarized domains, agreement is rarer, more meaningful, and more likely to shape voter inferences. Thus, the same general cue can vary in diagnostic value across issues, not because voters know which issues are involved, but because the structure of elite competition makes some cues more likely to reflect real movement.

In sum, the accuracy of the cooperation heuristic depends on the political environment in which it is applied. When elite constraints filter out symbolic cooperation, especially on polarized issues, perceptions of cooperation become more accurate indicators of policy proximity. When those constraints are weak, the same cue is less informative. This theoretical framework yields a clear prediction: perceived cooperation should more strongly predict policy proximity on highly polarized issues. ¹¹

Together, the discussion above specifies the conditions under which voters are likely to use inter-party cooperation as a policy heuristic. Each property—cheapness, simplicity, and accuracy—captures a distinct dimension of cue usability and is shaped by well-theorized moderators. Table 1 summarizes these three properties, outlines how each operates in the context of inter-party cooperation, identifies the relevant moderators, and presents the corresponding hypotheses. ¹²

OPEN IN VIEWER

Table 1.

Ecological Rationality Properties, Applications, and Hypotheses for Inter-Party Cooperation

Property	Definition (general)	Application to inter-party cooperation	Key moderators	Hypotheses: Relationship described in H1 will be stronger…
Cheapness	Ease of accessing or recalling the cue	Voters can readily access or remember information about cooperation between their preferred party and an out-party	Political knowledge	H2: Among politically knowledgeable individuals.
Simplicity	Ease of interpreting the cue once accessed	Cooperation is straightforward to interpret when the out-party’s ideology is distinctive	Ideological extremity of the out-party	H3: When the out-party is ideologically extreme.
Accuracy	Extent to which the cue reliably predicts the target outcome	Cooperation is more predictive of actual policy proximity when the issue is polarized	Issue polarization	H4: For polarized issues.

Research Design

To test the hypotheses above, I conducted original surveys in Canada (n = 1,000), Denmark (n = 1,278), Germany (n = 1,017), and the UK (n = 995) in 2019. ¹³ These four countries differ in party system and in their traditions of cooperation among parties, which allows for comparative assessments of how partisan cooperation shapes perceptions across distinct contexts. Because the meaning of cooperation between parties is conditioned by national experience, it is useful to situate each case within its respective tradition.

Canada’s political system is characterized by brokerage politics, where parties act as intermediaries between diverse social and regional interests (Carty & Cross, 2010;Cochrane, 2010). Rather than adhering to rigid ideological divides, parties frequently shift positions and engage in strategic cooperation to maximize electoral success (Johnston, 2010). Denmark, by contrast, operates within a multiparty system structured around stable “red” and “blue” blocs, where cooperation most often occurs within, rather than across, ideological camps (Green-pedersen & Thomsen, 2005). Governments are typically formed through coalitions of parties belonging to the same bloc, and cross-bloc cooperation, while possible, is less common and often politically costly. This tradition of bloc politics creates a context in which cooperation is expected and normalized within blocs but generates sharper signals when it occurs across them. Germany similarly has a long-standing tradition of coalition governance, including frequent grand coalitions between parties from different ideological camps such as the Christian Democrats and the Social Democrats (Debus et al., 2021). In this system, elite cooperation is frequent and institutionalized as a necessary feature of government formation. Finally, the UK represents a more adversarial case: a historically dominant two-party system where inter-party cooperation is rare and often perceived as a deviation from normal political competition. While coalition governments have occurred—most notably the Conservative–Liberal Democrat coalition from 2010 to 2015—they remain exceptional rather than routine (Barlow & Bale, 2021).

Thus, by spanning traditions that range from Canada’s brokerage politics to Denmark’s bloc competition, Germany’s institutionalized coalitions, and the UK’s adversarial two-party system, the study captures a wide spectrum of partisan cooperation. This variation ensures that the results are not artifacts of a single institutional or cultural setting. Instead, they provide evidence on how cooperation between a respondent’s preferred party and an out-party influences perceived policy proximity across contexts where cooperation between parties is routine, conditional, or unusual.

The next section describes how the key variables are measured. I begin with partisanship, since the analysis is limited to respondents who identify with a party — a restriction that follows directly from the theoretical framework. I then discuss the dependent variable (perceived policy distance across fifteen issues), the key independent variable (perceived cooperation between the respondent’s preferred party and an out-party), and the control variables. Descriptive statistics are provided in Appendix 2.

Policy Proximity

To measure the outcome variable of the paper, which is the proximity between partisan’s own policy position and the out-party’s position on that issue, I first need to identify which policies need to be included. The theory of ecologically rational heuristics suggests that cooperation cues are most likely to shape judgments on issues that are both salient in party competition and marked by partisan polarization. Guided by this logic, I include a broad set of specific policies: 15 in Germany and Denmark, and 14 in Canada and the UK. These policies were selected because they are central to contemporary political debates in each country and vary in their degree of partisan polarization.

While the specific content of elite discourse may evolve over time since political entrepreneurs often introduce new issues, reframe debates, or shift attention to different policy cleavages—the use of a “most-similar systems” design ensures a substantial degree of cross-national comparability in the relevancy of the policies. As such, the selected issues span a wide array of domains, including traditional economic policies, social and cultural issues, “new politics” topics, and country-specific policy issues.

In addition to the specific policy items, the survey includes a set of broader value-based questions designed to capture more abstract dimensions of political conflict that cut across policy domains. A long tradition of research suggests that enduring divides between parties often stem not only from disagreements over concrete policies, but also from deeper commitments to competing values—such as equality versus hierarchy or change versus tradition (Bobbio, 1996; Caprara & Vecchione, 2017; Jost, 2021). To reflect this dimension of political competition, the survey includes value-oriented items on income redistribution, cultural diversity, civil liberties, the role of government, and gender equality. Including both concrete policy stances and broader value commitments ensures that the dependent variable captures variation in both domain-specific preferences and underlying ideological orientations.

Across the four surveys, respondents are asked to evaluate all the selected issues by placing themselves and the major political parties on a common scale. Each issue is measured using a 1–7 scale, with the endpoints representing opposing positions on that specific policy or value dimension. The dependent variable is constructed as −1 multiplied by the absolute distance between a respondent’s own position and the perceived position of an out-party on each issue, such that higher values indicate greater perceived policy proximity. Table 2 lists all included issues, a brief description of the endpoints of each scale, and the set of political parties in each country that respondents were asked to evaluate. ¹⁵ Precise wordings of these questions are presented in Appendix 2. ¹⁶

OPEN IN VIEWER

Table 2.

List of Issues Included in the Surveys, Scale Endpoints, and Countries

Issues 17	Scale Endpoints (1 = …; 7 = …)	Countries
Tax and spend	Supports cutting taxes even if it leads to less public spending; opposes cutting taxes if it leads to less public spending	Canada, Denmark, Germany, UK
Tax and debt	Supports cutting taxes even if it leads to more public debt; opposes cutting taxes even if it leads to more public debt
Healthcare	Less to healthcare and pensions; more to healthcare and pensions
Immigration	Supports increasing immigration; opposes increasing immigration
Market regulation	Increase regulation of financial markets; decrease regulation of financial markets
Surveillance	Supports prohibition of surveillance; opposes prohibition of surveillance
Military	Less on military spending; more on military spending
Renewable energy	Supports taxing fossil fuels to promote renewable energy; opposes taxing fossil fuels to promote renewable energy
Income equality	Society works better when wealth is more equally distributed; society works better when those who are most productive get most of the wealth
Cultural diversity	Cultural diversity makes the nation stronger; Cultural diversity makes the nation weaker
Civil liberty	Protecting civil liberties is more important than improving national security; improving national security is more important than protecting civil liberties
Government’s role	Government should play a greater role in the management of the economy; government should play a smaller role in the management of the economy
Gender equality	There are legitimate reasons to treat men and women differently in economic and public life; men and women should be treated equally in economic and public life
Marijuana	Supports legal recreational use of marijuana; opposes legal recreational use of marijuana	Canada, Denmark, Germany
EU	Supports strengthening cooperation in the EU; opposes strengthening cooperation in the EU	Denmark, Germany
Brexit	Supports policies for a “hard Brexit”; opposes policies for a “hard Brexit”	UK

Note. The set of parties included are Canada: Liberal, Conservative, NDP, Bloc Québécois, Green, Germany: SPD, CDU/CSU, FDP, Greens, AfD, Die Linke. Denmark ¹⁸ : Social Democrats, Venstre, Red–Green Alliance, Danish People’s, Radikal Venstre, Socialist People’s, Liberal Alliance, Nye Borgerlige, Conservative People’s, Alternativet. United Kingdom: Labour, Conservatives, Liberal Democrats, SNP, Greens, UKIP.

Results

Table 3 presents the results most relevant to H1. Across the four countries, the relationship between respondents perceived cooperations and policy proximity is consistently positive and statistically significant, even after controlling for objective or subjective ideological distance between respondents’ preferred party and the out-parties, whether the two parties are perceived to be in government together, and policy fixed-effects. This means that when respondents perceive greater cooperation between their preferred party and an out-party, they tend to judge that out-party as being closer to their own position on aggregate across policy issues. ²¹ Substantively, a one standard deviation increases in perceived cooperation, which corresponds to 2.37 units in Canada, 2.78 in Denmark, 2.96 in Germany, and 2.53 in the UK, is associated with a 0.31, 0.34, 0.31, and 0.36 increased in perceived policy proximity on the 0–7 scale, respectively. To put these magnitudes in perspective, they are nearly as large as the effect of a one standard deviation decrease in perceived left–right distance—0.38 in Canada, 0.31 in Denmark, 0.33 in Germany, and 0.41 in the UK—suggesting that voters treat cooperation as an informational shortcut on par with the classic left–right heuristic. Furthermore, while the relationship holds across all four cases, the magnitude of the effect is somewhat larger in Canada and the UK than in Denmark and Germany. This may reflect differences in the informational value of cooperation across institutional contexts — where coalition governance is common, cooperation may more often reflect procedural necessity than genuine policy alignment, reducing its diagnostic value as a heuristic. Nevertheless, the fact that the effect remains clearly positive and significant even in Denmark and Germany underscores that the underlying relationship is not an artifact of any single party system or tradition of coalitional government.

OPEN IN VIEWER

Table 3.

Analyses of Policy Proximity

DV = policy proximity	Canada		Denmark		Germany		UK
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Perceived cooperation	0.1**	0.164**	0.1**	0.143**	0.09**	0.12**	0.11**	0.173**
	(0.005)	(0.005)	(0.005)	(0.005)	(0.004)	(0.004)	(0.004)	(0.004)
Manifesto LR distance	​	−0.017**	​	−0.0082*	​	−0.0032	​	−0.03**
	​	(0.004)	​	(0.0025)	​	(0.006)	​	(0.004)
Perceived LR distance	−0.142**	​	−0.112**	​	−0.128**	​	−0.15**	​
	(0.005)	​	(0.005)	​	(0.005)	​	(0.004)	​
Perceived cabinet membership	0.065**	0.098**	0.253**	0.275**	0.01	0.025	−0.019	0.046
	(0.025)	(0.025)	(0.03)	(0.03)	(0.03)	(0.027)	(0.025)	(0.026)
Constant	−1.97**	−2.41**	−2.26**	−2.67**	−2.24**	−2.7**	−2.31**	−2.61**
	(0.11)	(0.131)	(0.062)	(0.091)	(0.086)	(0.186)	(0.083)	(0.1)
Policy FE	Yes		Yes		Yes		Yes
Individuals	828		638		802		865
Observations	37,541		38,766		51,232		50,380

Regarding other variables, it is no surprise that perceived left–right distance between the in-party and the out-party is strongly associated with smaller policy proximity between the respondent and the out-party on a specific issue. The objective indicator from party manifestos also points in the same direction, reaching statistical significance in three of the four countries. Overall, this suggests that voters’ judgments of proximity tend to correspond more consistently with their own perceptions of ideological distance, while programmatic positions recorded in manifestos display a broadly comparable, though somewhat less uniform, association across contexts. A similar pattern emerges for cabinet membership: although the direction of the relationship is mostly positive, the relationship is only statistically meaningful in Canada and Denmark.

I also estimated several alternative models to assess the robustness of the findings. First, to account for the possibility that the association between perceived cooperation and policy proximity may vary depending on the strength of partisan attachment, I include a measure of respondents’ partisan strength. Second, I add a dummy variable indicating whether the out-party is a radical right party, given the possibility of radical right exceptionalism that may trigger distinct reactions (Gidron et al., 2023). Third, I replace perceived cabinet membership with an objective measure of whether the two parties are in government together. There are two variables capturing this objective measure: whether the two parties were in government together during the survey period (only applicable in Denmark and Germany), and whether they had record of joint governance in the last 10 years. ²² Furthermore, I substitute the subjective left–right distance between the respondent’s in-party and the out-party with the distance between the respondent’s self-placement and the out-party’s left–right position. This allows me to test whether the results depend on using a party-anchored measure of ideological distance or one that is anchored in respondents’ own self-placement. Finally, I address the possibility that missing responses to the party placement questions may bias the results. Because respondents could select a “Don’t Know” option, I re-estimate the models using stabilized inverse probability weighting (IPW) based on the predicted probability that respondents provide a substantive response (see Appendix 3.2 for details of the procedure).

All these alternative model specifications are presented in Appendix 3, and the substantive conclusion from Table 3 remains unchanged. Even after controlling for ideological distance between the in-party and out-party, their co-governance status, and policy fixed effects, the evidence continues to support H1: respondents perceive greater policy proximity to out-parties they see as cooperating with their in-party. This pattern is not just an artifact of issue selection. Even when holding constant all unobserved issue-level factors, respondents still see out-parties as closer to their own positions when they believe their in-party cooperates with them. To further assess this possibility, I conducted sensitivity analyses that sequentially exclude each issue from the analysis. Across the four countries, the estimated cooperation coefficient changes by no more than 2.6% to 4.9% when any single issue is removed. This stability suggests that the findings are not driven by any particular policy domain and remain robust across the set of issues examined (see Appendix 3.1).

Conclusion

This paper investigates partisan cooperation as a heuristic for judging policy proximity across a wide number of issues, making it one of the most comprehensive studies of its kind. Using original surveys from four multiparty democracies, I show that voters perceive out-parties as closer to their own views when they believe those parties are cooperating with their preferred party. This relationship holds even after accounting for perceived and objective ideological distance, shared cabinet membership, issue fixed effects, and unobserved heterogeneity across respondents and party-dyads. Moreover, the effect is shaped by conditions of ecological rationality: cooperation is more likely to be used as a heuristic when it is cheap (easily recalled by knowledgeable voters), simple (especially when the out-party is ideologically extreme), and accurate (particularly on polarized issues).

While these findings contribute to our understanding of how voters use elite behavior to make inferences, they also depart in important ways from existing research on coalition heuristics. Prior studies have shown that formal governing arrangements influence perceptions of ideological closeness between parties (Adams et al., 2016; Fortunato & Stevenson, 2013), but they often rely on institutional indicators and do not theorize the informational conditions under which such cues are likely to matter. This paper shifts the focus to voters’ perceptions of cooperation in order to capture informal, symbolic, and non-institutional signals, and develops a structured framework rooted in ecological rationality to explain when and why voters use cooperation as a policy heuristic. In doing so, it builds on but moves beyond earlier work that treats coalition status as a static cue for ideological proximity. Rather than assuming that voters uniformly rely on formal governing arrangements, this paper shows that perceived cooperation between one’s preferred party and an out-party, whether or not it reflects institutional power-sharing, can shape how voters judge the out-party’s proximity to their own position on specific policy issues. It also uses the ecological rational framework to theorize the conditions under which cooperation between parties serves as a useful heuristic cue, showing how elite behavior provides voters with signals that shape perceptions in complex political environments.

At the same time, this study has several limitations that point the way toward future research. The most important concerns causality. Because the analysis relies on observational survey data, the results should be interpreted as identifying associations between perceived cooperation and perceived policy proximity rather than definitive causal effects. Nevertheless, the ecological rationality framework does not require strict causality for the heuristic to be meaningful, so long as the cue reliably correlates with outcomes in the environment. Voters can learn through repeated exposure that cooperation and proximity tend to co-occur, making cooperation a useful mental shortcut even if cooperation itself does not produce policy convergence.

That said, experimental designs would be particularly valuable in adjudicating between alternative theoretical interpretations of why cooperation matters and in identifying whether exposure to cooperation cues causally shapes voters’ perceptions of policy proximity. Beyond the ecological rationality offered here, voters may respond to cooperation for different reasons: it may signal new partisan norms of civility, redefine partisan identities and group boundaries, or be taken as evidence of pragmatic power-sharing rather than ideological proximity. The current design also does not disentangle different types of cooperation. Citizens may react differently when cooperation involves explicit policy compromises compared to when it reflects non-policy gestures such as symbolic endorsements, joint appearances, or procedural coordination. Future experiments that vary the framing and content of cooperative acts would allow scholars to test whether these different forms of cooperation carry distinct informational value for voters. In this way, experimental approaches could clarify which dimensions of cooperation matter most, and why.

Beyond clarifying mechanisms and establishing causality, future work might also assess whether the cooperation heuristic outperforms alternative shortcuts. It would be particularly valuable to identify the target inferences for which cooperation provides more accurate judgments than other available cues, as well as the conditions under which its relative performance improves or deteriorates, and for which types of voters it is most informative.

Supplemental Material