Patterns of Legislative Agenda Setting: Evidence from Switzerland

2.1 Outcome of Interest

The basic prerequisite for any legislative project is that political decision makers pay attention to an issue. The issues that receive “serious consideration” from political decision makers constitute the institutional agenda (Cobb & Elder, 1971, p. 906). An issue can be on the institutional agenda for various reasons, such as focusing events, media reports, international actors and standards, and interest groups. However, an issue being on the institutional agenda by no means implies that a legislative process will follow. The literature therefore commonly uses the term legislative agenda for more narrowly defined agendas than the institutional agenda, that is, those related to actual legislation (Rasch & Tsebelis, 2011). In this paper, legislative agenda refers to those issues on which the administration received a formal legislative mandate.¹ Such a legislative mandate is the decisive starting point for the subsequent drafting process because such a mandate contains aims and policy alternatives for a draft law. This is the case for all laws, regardless of who was the decisive actor in legislative agenda setting (Ismayr, 2008; Müller & Uhlmann, 2013). Accordingly, this paper’s outcome of interest is an issue and policy alternative being put on the legislative agenda (LA).

In abstract terms, agenda setting is commonly understood as the process of transferring an issue from one agenda to an even narrower one (Green-Pedersen & Walgrave, 2014; Zahariadis, 2016). Therefore, in this paper, legislative agenda setting is understood as the process of transferring an issue from the institutional to the legislative agenda. To be precise, legislative agenda setting is a political decision-making process among political actors that negotiate whether and in what form a law should be drafted. Accordingly, in legislative agenda setting, issues and policy alternatives are selected and channeled into an institutionalized legislative process in which a first draft is produced. This means that the content of a future law is decisively foreshadowed in legislative agenda setting.

2.2 Introducing four Conditions

The debated question is which political actors are involved in legislative agenda setting and influence the legislative agenda, of course (Green-Pedersen & Walgrave, 2014). For example, authors discuss the power of the government (Rasch & Tsebelis, 2011) and parliament—the parliamentary committees in particular (Siefken & Rommetvedt, 2022)—in legislative agenda setting. The literature has identified the political actors that hold instruments to trigger legislative processes in Western European democracies, that is, to move issues lying on the institutional agenda closer to the legislative agenda. First, legislative projects can be launched within government and administration, and governments can eventually submit drafts to the parliament (Ismayr, 2008). Thus, the first condition to be considered for the analysis is active government with its administration (GA). Second, Western European parliaments have instruments to articulate their concerns for future legislative projects (Brunner, 2013).² Therefore, a second condition to be included for the analysis is active parliament (P). Third, in some Western European countries, (organized) citizens and subnational entities hold agenda-setting instruments. For example, initiatives that require a certain number of signatures from citizens to be launched in Switzerland, Italy, or Austria (Ismayr, 2008), and subnational entities can provide an impetus for legislative projects at the national level in Switzerland (Jaquet et al., 2019). Accordingly, the third and fourth condition to be considered for the analysis are active organized public (OP) and active subnational entities (SE).

2.3 Configurational Expectations

In line with the conceptualization of legislative agenda setting as a political decision-making process, an interplay is expected between the conditions named above. Generally, the literature assumes that the government/administration initiates a large share of all legislative projects (Schüttemeyer & Siefken, 2008). Olson and Norton (1996, p. 7) even formulated a “90 per cent rule,” according to which the government initiates 90 percent of the legislative projects. However, without doubting that the government/administration is an important player in the genesis of legislative projects (Bräuninger & Debus, 2009; Rasch & Tsebelis, 2011), the agenda-setting literature indicates that assigning a legislative project to one particular initiator is too simplistic (Seeberg, 2022). Consequently, it is implausible to claim that most legislative projects simply originate “from the government.” Rather, it is to be expected that legislative agenda setting consists of a plurality of impulses for legislative projects. For example, it is conceivable that the administration carries out preliminary work for a legislative project (e.g., publishes reports) and at the same time, parliamentary requests on the same issue are submitted.

Although the literature suggests that political actors interact in legislative agenda setting, it remains unclear which political actors’ activities and their interactions are consistently associated with the legislative agenda across cases. Most notably, the analysis below tests different expectations regarding the activities of parliament vis-à-vis the activities of the government with its administration. It is known that parliamentarians can influence policy making in any stages of policy making and that the government/administration and the parliamentarians may interact in any stages of policy making (Siefken & Rommetvedt, 2022). However, it is ambivalent whether the activities of parliament—and their interaction with the activities of government with its administration—are consistently associated with the legislative agenda across cases. On the one hand, the thesis of reparliamentarization claims that Western European parliaments have increased their power vis-à-vis the government in the past few years and decades (Goetz & Meyer-Sahling, 2008), and that parliamentary committees are powerful institutions with a lot of creativity and knowledge (Lüthi, 2009). Accordingly, one would expect that parliaments (parliamentary committees in particular) have increased their agenda-setting activity and are able to promote their input for legislative projects vis-à-vis the government/administration—at least in working parliaments. On the other hand, the thesis of deparlamentarization postulates that governments and their administrations enjoy increased power vis-à-vis parliament because of the increased complexity of social problems and a strong internationalization of legislation (Goetz & Meyer-Sahling, 2008). Accordingly, one would expect that the activities of the parliament are not consistently associated with the legislative agenda.

3.1 Fuzzy-set Qualitative Comparative Analysis

In line with the research question and conceptual/theoretical framework, which assume a potential interplay of various political actor’s activities, qualitative comparative analysis (QCA) was applied. This approach is based on the logic of set theory, meaning cases are treated as members/nonmembers in sets (conditions and outcome), and the approach allows for identification of necessary and sufficient conditions for a certain outcome (Schneider & Wagemann, 2012). At its core, QCA assumes causal complexity. Most notably, the approach suggests equifinality and conjunctural causation; that is, combinations of conditions (configurations) can lead to an outcome and different configurations can lead to the same outcome. The main idea of QCA is to systematically compare seemingly diverse cases to find concise cross-case patterns. More concretely, one follows a systematic minimization procedure to “filter” redundant conditions, that is, to depict conditions that are consistently associated with the outcome of interest across cases.

To be precise, the fuzzy-version of QCA—fsQCA—was applied, because the concepts allow for differences in kind and degree (Schneider & Wagemann, 2012). For example, it is not a binary question whether parliament is involved in legislative agenda setting; rather, parliament has various instruments that vary in terms of influence and difficulty of activation (Brüschweiler & Vatter, 2018).

3.2 Case Selection

A case was defined as the agenda-setting process of a legislative project. Therefore, a case ends with a legislative mandate for a draft law. To prevent “infinite regress back in time” (Princen, 2007, p. 23), at most, the first five years preceding a legislative mandate were considered. For the case selection, the political context was kept constant to ensure that the cases share enough background characteristics to be comparable (Berg-Schlosser & De Meur, 2009). Therefore, only agenda-setting processes in Swiss politics at the federal level were considered. Switzerland is an interesting laboratory for insights beyond Swiss politics because from a Western European cross-country perspective, many political actors hold institutionalized agenda-setting instruments and the parliament is considered an intermediary case in terms of its power vis-à-vis the government (Vatter, 2020). In addition, only legislative projects that were discussed in parliament between 2011 and 2021 were considered for case selection. This time period was chosen for practical reasons, that is, to ensure that enough high-quality data could be collected for all cases (see chapter 3.3 for more information about data).

Within these boundaries, in line with QCA as a diversity-oriented approach in which “a maximum of heterogeneity over a minimum number of cases” is recommended (Berg-Schlosser & De Meur, 2009, p. 21), cases were selected to cover various configurations (i.e., combinations of political actors’ activities). Which political actors are actively involved in a decision-making process depends heavily on the type of decision-making process: Europeanized, national, or federalist (Sciarini et al., 2015). Accordingly, three different policy fields were selected to cover the different types of decision-making processes: financial market (mainly Europeanized), migration (mainly national), and spatial planning (mainly federalist).³ In addition, when selecting the cases, it was ensured that multiple agenda-setting instruments (such as cantonal, popular, and parliamentary initiatives) were deployed.

Eventually, the sample consisted of 15 cases. This number of cases is large enough to conduct a QCA, thus, to find cross-case patterns. At the same time, this number of cases allows to gain within-case knowledge. For the selection of cases with a negative outcome (i.e., an issue is not on the legislative agenda), the “possibility principle” was taken into account: The outcome must be expected due to at least one condition; that is, at least one political actor must be actively involved in the process of legislative agenda setting (Mahoney & Goertz, 2004, p. 657). Since discussions and consultations within the government and the administration are confidential, cases in which a parliamentary instrument, a popular initiative, or a cantonal initiative was launched were considered as potential negative cases. See Table A1 in additional file 1 for an overview of the selected cases.

3.3 Data

Comprehensive and qualitative data were collected to acquire detailed case knowledge and substantively interpret QCA solutions. For one thing, about 30 semi-structured expert interviews were conducted and transcribed (see Table A2 in additional file 1 for an overview). Experts are people who were directly involved in the agenda-setting process or at least closely observed the events. For example, interviews with politicians, employees of the federal administration, parliamentary services (administration of parliament), and cantons were conducted. The interviews were conducted to obtain insider information on the decision-making processes in legislative agenda setting that may not be included in other data sources such as documents. For another, the database of the parliamentary services (“Curia Vista”), official documents of the authorities (such as reports of the federal administration, the explanatory report for the consultation, and the report of the consultation results), transcripts of the parliamentary committees’ meetings, parliamentary debates, media releases, media articles, and secondary literature were relied on. The data were triangulated; that is, data pieces were compared to validate the information.

3.4 Calibration

Finally, six-value fuzzy scales were used to calibrate the data, that is, to transform the data to membership scores between 0 and 1. Regarding the outcome (legislative agenda), the following were the qualitative anchor points. A score of 1 means that an issue is on the legislative agenda and that the legislative mandate includes very precise instructions on how to formulate a future draft. A score of 0.5 indicates that it is unclear whether an issue is on the legislative agenda. A score of 0 means that an issue is far from the legislative agenda (i.e., agenda-setting instruments such as parliamentary requests did not find majorities and fizzled out). Between 0.5 and 1, the level of detail of the legislative mandate for the subsequent drafting process was specified. For example, a fuzzy score of 0.6 means that a legislative mandate exists but does not specify any details for the drafting process. Between 0 and 0.5, it was graded how close an issue came to the legislative agenda. As an example of the calibration of the outcome, a score of 1 was assigned to case 13 (08.314) because the legislative mandate contained precise suggestions on how to formulate a future draft.

The conditions were calibrated as follows. A score of 1 means that a political actor was very actively involved in legislative agenda setting. A score of 0.5 implies that it is unclear whether a political actor was active. A score of 0 means that a political actor was not actively involved in legislative agenda setting at all. Between these three qualitative anchor points, it was differentiated how active the political actors were. For example, regarding the condition “active parliament,” a fuzzy score of 1 means that the Swiss parliament’s strongest instrument (parliamentary initiative) was submitted and that the initiative found a majority in the corresponding parliamentary committee. A fuzzy score of 0.8 was assigned when the parliament used less powerful instruments (e.g., motion) and formed a majority. A fuzzy score of 0.4 means that parliamentary requests and initiatives were submitted to parliament, but the parliament did not discuss them or did not form a majority in favor of the concern. As an example of the calibration of the condition “active parliament,” a score of 0.8 was assigned to case 6 (14.063) because a parliamentary committee mandated the administration to write a report on fundamental problems in the policy field of asylum and to propose policy alternatives. For the sake of transparency, Table 1 provides all fuzzy scores. To minimize the data matrix containing the calibrated data, the QCA and SetMethods packages in R were used. The final step in a QCA is to interpret the QCA solution with theoretical and in-depth case knowledge (also called dialogue with the cases).

4.1 Formal Analysis

The first step in a QCA is to test whether one of the conditions is necessary for the occurrence of the outcome, that is, whether the occurrence of the outcome is not possible without the presence of a certain condition. Considering the standard consistency threshold of 0.9 (Schneider & Wagemann, 2012), no necessary condition was found (see Table A3 in additional file 1). The condition “active government with its administration” (GA) comes closest to the consistency threshold, with a consistency of 0.806. Based on the insights from the reviewed literature and the theoretical framework, this finding can be interpreted as follows: Government/administration certainly play a role in legislative agenda setting, but they are not almighty and not the only players in legislative agenda setting.

The second step is truth table analysis. We are now concerned with sufficient configurations, that is, combinations of conditions that are always present when the outcome occurs. As recommended (Schneider & Wagemann, 2012), large gaps between truth table rows in terms of consistency and PRI were considered for the decision regarding the consistency threshold, and therefore chose a consistency threshold of 0.8 was chosen. The truth table (see Table A4 in additional file 1) shows that empirical cases cover 9 of the 16 possible configurations, which is remarkable. The truth table can be interpreted as follows: The government, with its administration, is severely constrained in legislative agenda setting. First, all consistent configurations involve multiple active political actors. Second, the consistent configurations show that whenever the government, with its administration, is active, other political actors are, too. Third, one consistent configuration shows that it is possible that the government with its administration is not active but other political actors are.

Step three is the systematic minimization of the truth table, which results in a solution formula (see Table 2). As Schneider and Wagemann (2012) suggested, the intermediate solution was chosen for the interpretation. The corresponding solution has a remarkable consistency, 0.867, and a high coverage, 0.722, so the statement that the solution is sufficient is true to a very large degree, and the solution explains most of the cases. To be precise, the solution formula contains three paths. Path 1 consists of the combination of “active government with its administration” and “active parliament” (GA*P). Paths 2 and 3 comprise “active organized public” in combination with “active government with its administration” (GA*OP) or in combination with “active parliament” (P*OP), respectively (see additional file 1 for some comments on the robustness of the fsQCA solution).

4.2 Dialogue with the Cases

The final step is a dialogue with the cases, which means that the solution formula is interpreted using case knowledge. Paths 2 and 3 are straightforward: The organized public, in the analyzed cases interest groups, launched popular initiatives in response to the government’s and parliamentary majority’s intentions. The government and/or parliament, in turn, tried to fend off these initiatives with counterproposals. For example, a person from the administration said in an interview about case 14.023, “That [the popular initiative] was not what we wanted.”

Path 1, consisting of “active government with its administration” and “active parliament,” not only has much higher coverage scores but is also much more interesting from an analytical point of view. The interpretation and discussion will therefore focus on path 1. The dialogue with the cases suggests that path 1 is a robust finding and further reveals the following. In all empirical cases path 1 covers, “parliament” more precisely means “parliamentary committee.” The cases are characterized by the fact that government, with its administration, and a parliamentary committee want to put an issue on the legislative agenda, but the parliamentary committee articulates and prefers policy alternatives different from those of the government and its administration. In other words, various ideas are present regarding the direction of a future draft. Therefore, the parliamentary committees successfully impose their preferred policy alternatives on the government with its administration, so the parliamentary committees decisively shape the legislative agenda in an interaction with the government/administration. For example, in case 14.063, the government/administration submitted a relatively minor revision of the asylum act to parliament. Their aim was to accelerate asylum procedures. However, the parliamentary committee criticized that this goal can hardly be achieved with the proposed revision and mandated the administration to write a more detailed report about asylum procedures with new policy alternatives. A person from the administration said in an interview, “The parliamentary committee told us to start from scratch again.”

Parliamentarians/parliamentary committees have several possibilities to participate in legislative agenda setting. First, in Swiss politics, so called legislative projects under a parliamentary committee’s lead exist. This means that a parliamentary committee can form a majority for a cantonal initiative (case 08.314) or parliamentary initiative (case 16.403) and initiate a legislative process against the will of the government/administration. For example, a person from the administration said in an interview about case 08.314, “We promoted different policy alternatives and recommended to vote against the cantonal initiative.” Second, case 14.063 shows that parliamentary committees may also use the deliberations of government bills for legislative agenda setting. Third, it is (theoretically) possible that parliamentarians/parliamentary committees launch motions that the administration has to implement or engage in informal exchanges with the administration.