Introduction
Social scientists researching bicameralism and the influence of upper chambers on the functioning of parliaments have traditionally focused on two main areas. Firstly, considerable attention has naturally been devoted to the constitutional delineation of the upper chambers, primarily examining the powers of upper chambers in relation to lower chambers (Lijphart, Reference Lijphart1984; Reference Lijphart2012; Sartori, Reference Sartori1994; Siaroff, Reference Siaroff2003). The veto power of the upper chamber plays a crucial role in this regard (Riker, Reference Riker1992; Tsebelis, Reference Tsebelis2002), as well as the nature of other powers, such as the authority of the upper chamber in relation to the government. Secondly, it is equally important to consider the composition of the upper chamber in comparison to the lower chamber. The form of representation in both chambers, taking into account congruence or incongruence in terms of identical or differing prevailing interests, also significantly influences the shape of bicameralism (Rogers, Reference Rogers2001; Druckman and Thies, Reference Druckman and Thies2002; Lijphart, Reference Lijphart2012; Giannetti et al., Reference Giannetti, Pedrazzani and Pinto2020; Pedrazzani and Zucchini, Reference Pedrazzani and Zucchini2020; Kim et al., Reference Kim, Hicken and Kollman2024).
Unlike other common constitutional institutions within political systems, upper chambers do not have a guaranteed place in the system. There are many systems with only one chamber – unicameralism. Some states (for instance, Denmark, Croatia or Sweden) have even decided to abolish their upper chambers of parliament (Nilsson, Reference Nilsson, Bijleveld, Grittner, Smith and Verstegen2020, 143–44; Skjæveland, Reference Skjæveland, Bijleveld, Grittner, Smith and Verstegen2020; Zrile, Reference Zrile2021), and in others (e.g. Canada or Ireland) this step is being contemplated (MacCarthaigh and Martin, Reference MacCarthaigh and Martin2015; Rancourt et al., Reference Rancourt2023). Upper chambers potentially challenge majority rule and they may also suffer from a democratic deficit (Russell, Reference Russell2001b; Reference Russell2013). Alternatively, they can simply be perceived as unnecessary. In this regard, it is logical that legitimacy is a concept closely associated with the upper chambers and bicameralism in general.
Although Lijphart (Reference Lijphart1984; Reference Lijphart2012) and Money and Tsebelis (Reference Money and Tsebelis1992) have highlighted the significance of legitimacy in their studies of upper chambers, they do not extensively develop this theme. The importance of legitimacy in analysing bicameralism was more explicitly developed by Russell (Reference Russell2013). In her influential article analysing the United Kingdom, Canada, and Australia, she identified perceived legitimacy as the third dimension of bicameralism. Russell explained that a lack of legitimacy undermines the ability of the upper chamber to exercise its institutional powers, while strong legitimacy can be utilized to engage in the political process and balance the lack of formal powers. Therefore, legitimacy complements the other two dimensions of bicameralism (formal powers, and compositional incongruence) but does not replace them (see Figure 1).
Following Russell’s work, other authors have delved into the topic of the legitimacy of upper chambers (Medeiros et al., Reference Medeiros, Bol and Nadeau2018; Mueller et al., Reference Mueller, Vatter and Dick2023; Štrus and Brezovnik, Reference Štrus and Brezovnik2023; Vercesi, Reference Vercesi2017). Mueller et al. (Reference Mueller, Vatter and Dick2023), following Vercesi (Reference Vercesi2017) – the author of the diagram shown in Figure 1 – sought to develop a new bicameralism index that incorporates the legitimacy of upper chambers, alongside the two previously considered dimensions. This follows Russell’s thesis and enables a more comprehensive comparison of upper chambers. However, these attempts have faced complications due to the still inadequately explored concept of the legitimacy of upper chambers.
Russell (Reference Russell2013) describes the concept as perceived legitimacy. Then the question arises: perceived by whom? In one part of her text, Russell (Reference Russell2013: 375) asks, ‘in what circumstances will a second chamber be perceived as legitimate by the public and elites?’ Similarly, in the context of perceived legitimacy, Mueller et al. (Reference Mueller, Vatter and Dick2023: 328) mention ‘citizens and/or elite members’. Should the focus thus be on the perceptions of citizens or elites, especially the members of the upper chamber themselves?
Another problem in previous attempts to measure the legitimacy of upper chambers is the excessive emphasis on how upper chambers are formed, particularly on the level of representativeness and democratic legitimacy gained through direct elections (Lijphart, Reference Lijphart2012: 193; Vercesi, Reference Vercesi2017: 612). Conversely, other aspects of legitimacy have often been overlooked. In this regard, the importance of various components of the legitimacy concept beyond input legitimacy remains mostly unclear.
Despite offering probably the most advanced assessment of bicameralism considering the legitimacy dimension, Mueller et al. are aware of the limitations of their study. Hence, they repeatedly state that ‘ideally, they would dispose of citizen and/or elite surveys telling us what is thought of second (upper) chamber legitimacy’ (2023: 316). This approach would enhance our comprehension of the significance of the individual components within the concept of the perceived legitimacy of upper chambers. Consequently, it would enable effective conceptualization of perceived legitimacy in future comparative studies. That is precisely the subject of this study.
The research is based on direct, in-depth interviews with both the general public and elites (members of the upper chamber) in order to (1) explore different aspects (beyond frequently addressed democratic legitimacy) of the upper chambers’ perceived legitimacy and their significance and thus provide a foundation for better conceptualization, and (2) to explore whether there are significant differences in perception between the general public and elites, identify possible reasons for any such difference and thus enable discussion on whose perception is more important for the actual performance of the institution. This study thus adopts an exploratory approach, aiming to generate conceptual insights rather than to test predetermined hypotheses. It is intended not to be in opposition to quantitative attempts to measure the legitimacy of upper chambers, but, on the contrary, to contribute to a more precise conceptualization for future comparative studies. It should thus be seen as another piece of the puzzle in exploring the legitimacy of upper chambers.
The case of Czech bicameralism is well suited for examination for several reasons. As the problems that Mueller et al. while operationalizing the Czech case suggest (2023: 321–22), the Czech Senate offers a good illustration of how previous attempts to measure the legitimacy of upper chambers have foundered due to an inadequate conceptualization and operationalization of the concept. The Czech Senate is directly elected, but elections are characterized by an extremely low voter turnout. That raises the question of how to understand its input legitimacy. This case also illustrates very well the problem of overemphasizing input legitimacy.
The legitimacy of upper chambers in this landscape – especially in the context of democratic backsliding that has occurred in the region of East-Central Europe (ECE) in recent years (see Bakke and Sitter, Reference Bakke and Sitter2022; Castaldo and Memoli, Reference Castaldo and Memoli2024; Hanley and Vachudova, 2019) – may thus have a greater bearing on the preservation of democratic stability than in established Western democracies. In other words, the upper chambers in ECE countries are being tested as to whether they can fulfil one of their fundamental roles – to ensure the stability of the (democratic) system. Their perceived legitimacy is thus very important as they potentially challenge the majority rule. And, as Maria Romaniello emphasizes, upper chambers can only fulfil their functions of reflection and control if they are perceived as legitimate (Reference Romaniello2019). We also should not forget that, compared to Western Europe, bicameralism in Eastern European countries is historically under-researched in general. The same applies for non-federal bicameralism.
Legitimacy of upper chambers
Regarding the study of institutional legitimacy, I adopt an empirical approach similar to Russell (Reference Russell2013: 374–75) rather than viewing legitimacy as a sociological phenomenon (Lipset, Reference Lipset1959; Weber, Reference Weber and Parsons1947). Russell (Reference Russell2013: 375), referring to Mill (Reference Mill1861/1998) and Mezey (Reference Mezey1979), argues that social support is a crucial determinant of the behaviour of the upper chamber. ‘If it has support, it is more likely to demonstrate the confidence to challenge the government, and subsequently, to have its interventions taken seriously’, claims Russell (Reference Russell2013: 375). In this context, Russell also cites Dogan’s (Reference Dogan, Chehabi, Stepan, Kasza, Lijphart and Dogan1995) definition, stating that if ‘people hold the opinion that existing institutions are appropriate or morally proper, then those institutions are legitimate’. This clearly refers to legitimacy as perceived by the general public. However, in the end, it is the members of the upper chamber who decide whether to take or not to take a particular action. Hence, their perceptions of legitimacy are at least equally significant, if not more so. Senators’ perceptions may reflect the opinions of the general public, as senators are also part of society. On the other hand, it is possible that a certain gap in perceptions of legitimacy exists between the general public (i.e. ordinary citizens) and senators.
This gap in perceptions between elites and the public is not uncommon (e.g. Dellmuth et al., Reference Dellmuth2022, 188–212). It holds true even in the context of the Czech Republic (see, e.g. Kalhousová et al., Reference Kalhousová2025) – even though it has not been documented specifically in relation to perceptions of the role and legitimacy of political institutions. The elite-public gap can stem from differences in knowledge and experience (Hafner-Burton et al., Reference Hafner-Burton, Hughes and Victor2013; Riker, Reference Riker1995), variation in personality traits relevant to political engagement (Dynes et al., Reference Dynes, Hassell and Miles2019; Linde and Vis, Reference Linde and Vis2017), or positive selection mechanisms favouring more competent individuals (Dal B’o et al., Reference B’o2017). On the other hand, Kertzer’s meta-analysis (Reference Kertzer2022) tends to mitigate the perceived magnitude of the elite-public gap.
As I have already indicated, the legitimacy of the upper chamber should be understood more broadly than merely in terms of democratic legitimacy related to direct elections (see Lijphart, Reference Lijphart2012: 193) or the degree of representation of minorities (e.g. Lijphart 1991) and/or regional entities (see Vercesi, Reference Vercesi2017). The original work of Russell (Reference Russell2013: 375–76) therefore employs the concept of institutional legitimacy, as previously developed and utilized by various authors (e.g. Bodansky, Reference Bodansky1999; Papadopoulos and Warin, Reference Papadopoulos and Warin2007; Scharpf, Reference Scharpf1999; Schmidt, Reference Schmidt2013; Warren, Reference Warren2017), and as further employed by Mueller et al. (Reference Mueller, Vatter and Dick2023). Within this conceptualization, institutional legitimacy is understood in a much more complex manner and consists of three components: input (or source), throughput (or procedural), and output (or substantive) legitimacy.
-
1) Input Legitimacy – Input legitimacy refers to the source from which an institution derives its legitimacy. In the case of democratic regimes, this source is the people. Therefore, Mueller et al. (Reference Mueller, Vatter and Dick2023: 315–16) argue that the shorter the delegation chain from citizens to decisions, the greater the input legitimacy. Input legitimacy thus pertains to the manner in which the upper chamber is constituted. However, as Russell (Reference Russell2013) points out, for upper chambers, this involves not only democratic legitimacy in terms of direct elections (and associated voter turnout) but also representativeness in terms of representation of those whom the institution is intended to represent. In practice, for upper chambers, this can involve the representation of territorial interests (Russell, Reference Russell2001a) or specific minorities (e.g. Couwenberg, Reference Couwenberg, Hans, Blockmans and de Schepper1992; Russell, Reference Russell2001b: 443–46). Input legitimacy can also be based on the expertise, wisdom, or other qualities of the members of the institutionFootnote 1 (Blom, Reference Blom, Hans, Blockmans and Schepper1992; Vibert, Reference Vibert2007) or on the capacity of members to offer perspectives different from those in the lower chamber (Lijphart, 1991; Lipset, Reference Lipset1959; Mughan and Patterson, 1999; Russell, Reference Russell2001b). Hence, directly elected chambers may not necessarily be seen as legitimate, as Russell (Reference Russell2013: 383–85) illustrates in the case of Australian bicameralism, and conversely, indirectly elected or constituted chambers (or chambers elected in elections with very low voter turnout) may not necessarily be deemed less legitimate. Farrington has described this as non-electoral legitimacy (Reference Farrington2015).
-
2) Throughput Legitimacy – Throughput legitimacy encompasses the ‘quality of the governance processes as established by their efficacy, accountability, transparency, inclusiveness and openness to interest intermediation’ (Schmidt, Reference Schmidt2013: 6). In other words, throughput legitimacy refers to the quality of the processes by which the upper chamber is governed. In this regard, Russell (Reference Russell2013: 376) highlights, using the example of courts, that processes perceived as less political may be considered more legitimate (see Baird, Reference Baird2001). This can be particularly true in a context where there is sufficient demand in society for anti-politics (Hruška and Balík, Reference Hruška and Balík2024: 15–16). It is also crucial to note that strong procedural legitimacy can theoretically compensate for a deficit of input legitimacy (Hibbing and Theiss-Morse, Reference Hibbing and Theiss-Morse1995). This is a significant observation in the context of upper chambers, which are known for being less partisan and, consequently, more independent (and theoretically more thorough) in their oversight functions (Russell, Reference Russell2001b, Reference Russell2013). Moreover, the processes of upper chambers are often less covered in the media (Russell, Reference Russell2001b: 452), which can influence their perceived legitimacy.
-
3) Output Legitimacy – Output legitimacy refers to general satisfaction with the institution’s demonstrated performance and, consequently, specific political outcomes. As it involves perceived legitimacy, it is also based on perceived results. Therefore, what matters is not always the actual outputs of the upper chamber but rather how relevant actors perceive them. It is important to note here that the work, results, and overall contributions of upper chambers may not be well-known (Hruška, Reference Hruška2023). That can theoretically significantly affect the perception of upper chambers’ performance. Similar to throughput legitimacy, in regard to the popularity of upper chambers’ interventions, output legitimacy can compensate for a lack of input legitimacy (Menon and Weatherill, Reference Menon and Weatherill2008; Russell, Reference Russell2013).
Nevertheless, the question remains regarding the extent of the impact of each component on overall legitimacy. I believe it is vital to consider all components – something previous studies have usually neglected. Yet, it is unclear whether all components of the concept have equal influence. Does legitimacy depend equally, for instance, on the transparency of processes as on the fulfilment of institutional expectations? To what extent can one type of legitimacy compensate for another in the case of bicameralism? Moreover, does the importance of individual legitimacy components differ between the general public and elites?
Current research also seems to suffer from an oversimplification of the concept due to the need for viable operationalization. Mueller et al. (Reference Mueller, Vatter and Dick2023) rely, among other methods, on expert questioning. However, their approach does not consider the perceived legitimacy of the actors within the system but the legitimacy as perceived by experts on parliamentary deliberation, which can substantially differ. Mueller et al. operationalize the output legitimacy of upper chambers by analysing the occurrence of attempts at major reforms of upper chambers and attempts to abolish them. However, these attempts may result not from a lack of perceived legitimacy (by citizens or the members of the upper chamber) but could be, for instance, part of the political strategy of specific actors (Vercesi, Reference Vercesi2022) or an illiberal development (e.g., see Havlík, Reference Havlík2019: 380; Petrúšek and Kudrnáč, Reference Petrúšek and Kudrnáč2019: 69). To address described challenges, an analysis of direct interviews with both types of actors is employed.
Bicameralism in the Czech context
The Czech Senate is a relatively new institution; it was established as part of the constitutional order of the Czech Republic. Although the Constitution envisaged the existence of the Senate from the state’s very foundation (January 1, 1993), the institution itself was, in fact, not created until the end of 1996. According to Kopeček (Reference Kopeček, Balík, Hloušek, Kopeček, Holzer, Pšeja and Roberts2017: 126–29), the stigma of being unnecessary, together with persistent and heavy criticism from some politicians over time,Footnote 2 have influenced the public’s scepticism of the institution.
Senators are directly elected for six years in single-member constituencies using a two-round majority system, with the top two candidates from the first round moving to the second round. If a candidate receives a majority of votes in the first round, there is no second round. This, however, rarely happens. The Senate has 81 members and is continuously renewed in thirds; 27 senators are elected every two years. There is a persistent pattern of very little public interest in the Senate elections. The average turnout for the first round of the Senate elections is 37%, and for the second round, only 21.7% (Volby.cz, Reference Volby2024. Czech Statistical Office).
Czech bicameralism is asymmetrical. The Senate participates in the legislative process but can be overruled by a majority of all members of the lower chamber; it has only a suspensive veto of actions. However, the Senate’s consent is required in some other specific matters, including the approval of constitutional and electoral laws, the ratification of international treaties, the declaration of various kinds of emergencies, and the deployment of the Czech armed forces outside the Czech Republic. The Senate also has some other, mainly creation or control, powers. Among other things, it approves the appointment of constitutional judges. It also has the legislative initiative. According to Patterson and Mughan (Reference Patterson and Mughan2001), the Czech Senate is thus an upper chamber with delay and advisory functions.
The Czech Senate is an upper chamber founded on the principle of horizontal separation of powers (Loewenstein, Reference Loewenstein1969: 167–70). It is not formally designed to represent territorial interests (Hruška and Balík, Reference Hruška and Balík2024), although some individual senators do take on this role in practice (see Hruška and Čapek, Reference Hruška and Čapek2025). While many upper chambers are explicitly structured to represent specific interests – particularly territorial ones in federal or decentralized unitary states – chambers based on the principle of legislative duplication are equally common. Haas (Reference Haas, Riescher, Ruß and Haas2010) thus distinguishes between two fundamental roles of second chambers: (1) the principle of representation, which refers to the articulation of particular interests, and (2) the functional principle, which relates to the chamber’s role in the separation of powers and legislative oversight. Similarly, Tsebelis and Money (Reference Tsebelis and Money1997) differentiate between a political argument – referring to the representation of distinct societal interests – and an efficiency argument, focused on legislative oversight and the technical refinement of laws. It is, however, common for upper chambers grounded in interest representation to also perform oversight functions. Importantly, many upper chambers, regardless of their institutional foundation, face similar challenges related to their perceived legitimacy (Coakley, Reference Coakley2014; Russell, Reference Russell2001b).Footnote 3
Data and methods
Since I aim to explain how the different components of the legitimacy concept shape perceived legitimacy, I employ qualitative content analysis of semi-structured interviews, which are well suited for capturing and understanding the specific perceptions of different actors (see, e.g. Rubin and Rubin, Reference Rubin and Rubin2011). Also, political science is seeing a reviving interest in the study of parliaments via one-on-one interviews with legislators (see Steiman and Suhay, Reference Steiman and Suhay2023). This study thus continues this trend.
In order to capture potential differences in the perceptions of citizens and members of the institution, this study is based on two rounds of semi-structured interviews. First, I interviewed 30 citizens of the Czech Republic in November and December 2021. To find and select respondents, I advertized the research project in two ways. The first involved sharing the call on various online platforms and social media networks. Prospective participants who expressed interest were required to complete a brief online questionnaire. This questionnaire sought information on their socioeconomic characteristics and voting behaviour. Recognizing the need to include individuals not actively engaged in online spaces, such as the elderly, I also asked acquaintances to invite people from their personal networks. Over 500 respondents volunteered to participate through these two methods. From this pool of volunteers, 30 respondents were selected for the initial round of interviews. This number was estimated to be sufficient to achieve saturation.Footnote 4 The project budget allowed for an additional round of interviews if saturation was not reached, however, continuing in interviewing beyond 30 interviews proved unnecessary, as saturation was successfully attained.
An exception to random selection was made in an attempt to create a sample that was diverse in the basic socioeconomic characteristics of the respondentsFootnote 5 . Deliberate efforts were made to ensure heterogeneity in variables such as gender, age, size of place of residence, highest educational attainment, and personal income. At the same time, the sample consisted of two subgroups of respondents on the voting scale – those who regularly voted in all elections, including to the Senate (15), and those less willing to participate in Senate elections compared to other elections (15). Both subgroups had similar socioeconomic characteristicsFootnote 6 . In this way, I tried to ensure that I capture perceptions representing different parts of society.
To learn about senators’ perceptions, 14 current senators (out of a total of 81 members of the upper chamber) were interviewed between April and July 2023. As with the ordinary citizens, I sought to achieve a diverse sample. Thus, the following characteristics factored into the selection process: political affiliation, experience in the Senate, partisan vs. nonpartisan senators, backbenchers vs. members of the Senate presidency or committee Chairs, and also the types of constituencies – (1) rural vs. urban, and (2) those with natural borders corresponding to administrative units, that is, statutorily defined counties vs. constituencies with artificial borders. Again, the aim was not to create a fully representative sample, but to include senators with various characteristics that could potentially influence their responses.
I received a predominantly positive response when reaching out to senators and their assistants. Although I was turned down in a few cases, I was able to approach other senators with the same or similar characteristics. The fact that I could relatively smoothly secure access to senators for face-to-face meetings can be considered as one of the strengths of the researchFootnote 7 . Again, the strategy did not involve setting a predetermined number of senators to interview; instead, I aimed to continue the interviews until reaching the point of saturation. As with the interactions with the general public earlier in the process, the final interviews brought no new insights. The point of saturation was reached when conducting more interviews no longer served any purposeFootnote 8 .
In the interviews with both groups, I did not talk directly about legitimacy from the beginning. Instead, I first asked about issues related to legitimacy (e.g. the perceived performance of the institution, the characteristics of the senators etc.) and the different dimensions of the legitimacy concept as described in the theoretical section above. Only after that did I ask directly about the legitimacy of the Senate.
All the interviews were recorded. The data were subsequently subjected to qualitative content analysis. The research logic can be described as abductive (see, e.g. Tavory and Timmermans, Reference Tavory and Timmermans2014). It means that in the responses of citizens and senators, we were primarily looking for codes referring to existing theory; in this case, the different types of perceived legitimacy, that is, input, throughput, and output legitimacy and their specific manifestations (e.g. anti-political deliberation as part of throughput legitimacy). At the same time, however, we remained attentive to possible new factors not predicted by theory which could also influence perceptions of the Senate’s legitimacy (e.g. effect of dual mandates) (see Table 1). Subsequently, we focused on the contexts in which the codes occurred, that is, which specific factors had positive and negative impacts on the perceived legitimacy of the upper chamber, and which particular codes were essential for the overall legitimacy of the institution within both groups of respondents. We also compared these factors between the two groups. In order to strengthen the reliability of the results, coding was performed independently by two scholars. There was no significant coding inconsistency – the intercoder agreement value was 93.1%.
Table 1. Categories of main codes
Analysis and results
Input legitimacy
Observing the extremely low turnout of the Czech Senate elections, Mueller et al. reduce the input legitimacy of the Czech Senate in their index (Reference Mueller, Vatter and Dick2023: 321–22). However, the perception of senators differs. Senators included in the sample mostly did not perceive the low turnout as something that lowers the input legitimacy of their institution. They often stressed that although turnout is low, the elections are free and fair and therefore valid. The decision of citizens not to vote was perceived as their free decision, which does not affect the institution’s legitimacy. A typical answer is that of senator no. 5: ‘Everyone has the right to vote, not to vote’. or of senator no. 3: ‘I think not [the Senate does not have a legitimacy problem] because, simply, it is elected’.
The senators also pointed out that low voter turnout is a problem primarily in the second round of the electionsFootnote 9 . They again referred primarily to practical procedural and technical arguments that they felt were not very related to input legitimacyFootnote 10 . In the same context, the senators often stressed that the repeated criticism of the Senate’s legitimacy comes from populists and those who have consistently failed in Senate elections.
For ordinary citizens in the sample, low turnout is a more important issue. However, as I will show later, the legitimacy of the Senate from the citizens’ perspective is primarily related to output legitimacy, and therefore, the overall issue of input legitimacy (including turnout) was not articulated as often as it would be expected based on the existing theory. Moreover, low turnout was perceived by interviewed ordinary citizens as a consequence of low Senate legitimacy rather than a reason for it.
From the senators’ perspective, an aspect that greatly reinforces the Senate’s input legitimacy is the electoral process, whereby one-third of the Senate is elected every two years. The testimony of senator no. 10 illustrates this well:
There were a few times in the past when a populist movement shone, and after a while, it fell again. So, it is possible to get political power in a certain period of time and to get influence in the lower chamber. However, with the Senate, it is more complicated … because the Senators are elected by thirds. So if populists want to win the upper chamber and have the constitutional majority, they cannot do it in one electoral year but they need six years. And that’s long enough for people to wake up and realise that it is going in the wrong direction.
The specific method of continuous elections is thus seen as a kind of guarantee of stability and a barrier against extremism and populism. The type of electoral system may therefore be another aspect that can influence the legitimacy of an institution.
However, this only applies to the perceptions of senators. Most of the ordinary citizens in the sample were not aware of this important aspect of input legitimacy. The only exceptions were several respondents who demonstrated a well above-average level of knowledge of the Senate during interviews. This difference in perception between senators and the general public is probably an effect of very low knowledge about the Senate, with most people unfamiliar with the Senate’s electoral mechanism (see Hruška, Reference Hruška2023). The gradual change makes the Senate elections rather chaotic for many interviewed citizens, thus undermining input legitimacy.
An important feature of input legitimacy is also the connection of individual senators to their constituencies. Even if they are not legally designated to do so, senators often adopt the role of local ombudsmen, frequently also holding a mandate in local politics. From the perspective of some senators, this is one of the main, if not the main, sources of their legitimacy. Senator no. 7 expressed it well:
I wanted to end my career as mayor [prior to running for the Senate], which I had been basically since the 90s. And the fellow mayors said, ‘no, we need you because that would be a waste not to use you, okay?’ … And then, the mayors decided they wanted to have a representative in the Senate. … There are 81 constituencies, 81 winners. It has the advantage that people come to you and tell you about their problems.
Thus, for senators, the direct connection to their constituents and less distance from the citizens strengthens their input legitimacy. Related to this is a certain micromanagement that is typical of many senators in the sample. ‘I confess that as a senator, I can help a lot more those people, all those companies, and the municipalities that contact me because I have more time to spend with them’, said senator 11, explaining why they prefer being senator over member of the lower chamber. But it should be mentioned that not all senators see their role in the Senate as representing individual constituencies.
In this respect, the perceptions of interviewed senators and citizens did not differ much. For ordinary citizens, the direct link to their senators and their advocacy for the constituency’s interests positively impacted the institution’s perceived legitimacy. However, this was the case only for citizens who knew about this potential role; only some citizens in the sample perceived the Senate as an institution representing the interests of individual constituencies.
Another aspect that, for several senators in the sample, strengthened input legitimacy, and which is often related to representing local interests, was closer contact with the public. Senator no. 1 described it as follows:
The Senate really has a big role in that it also offers a space for debate with civil society…. The Senate is very active in holding all sorts of public hearings and conferences that often precede that legislation’s formulation. I think it is actually not really a visible role of the Senate but quite an important one.
On the other hand, several senators criticized the lack of communication with citizens and the public in general.
Senators also often associated input legitimacy with their own personalities, and not with the Senate as an institution. This is because, notwithstanding low turnout, winning a highly personalized election gives senators strong input legitimacy, from their perspective. This is evidenced, among other things, by the fact that, when interviewed, almost all the senators proudly stated that the Senate is a ‘chamber of winners’. In addition, the senators viewed the Senate as an institution whose input legitimacy was strengthened because of its members’ expertize, wisdom, and other personal qualities. The dominant view of the personal composition of the Senate is well captured by senator no. 4’s statement: ‘They have more life experience behind them…. And they are, I would say, prudent, they stay on top of things’, and senator no. 10: ‘These are people who have had to succeed in their professions to become successful and well-known and thus electable, so I guess they must be purposeful, ambitious’.
The vast majority of senators expressed similar views, and the perception of senators in the sample was consistent. However, the general public’s perception differed and was far from consistent. Some interviewed ordinary citizens shared the perception of the senators. However, according to another group of ordinary citizens in the sample, the Senate is perceived as a dumping ground for retired or unsuccessful politicians, where mainly older politicians go for a reward or compensation (similar to the filling of positions on various supervisory boards). According to respondent no. 4, whose statement is telling in this context, politicians who ‘didn’t make it to the Chamber of Deputies, weren’t elected, often go…I don’t want to say compensation, but you can use that too…they are just moved to that Senate’.
The Senate’s legitimacy may also be perceived negatively because the institution is not very representative in the sense of reflecting the structure of society. Particularly among liberal-minded citizens, the Senate may be perceived as too conservative with regard to its age and gender composition (‘the old men’s chamber’). However, at least in this study, this was very much a minority view.
Throughput legitimacy
Throughput legitimacy plays a role in the perceptions of both senators and the general public, but not in the same way. Interviewed senators very often emphasized that the Senate’s culture is different from that of the lower chamber. They described the Senate’s culture as a sophisticated, decorous workplace where debates are factual and focused on the specific topic under discussion. Compared to the Chamber of Deputies, the Senate does not, in the senators’ view, see fierce verbal battles, including flaming emotions, altercations etc. Overall, according to the senators interviewed, there is less ‘drama’ in the Senate. Senator no. 4 described it as follows: ‘We here, in the Senate, are also very concerned about the correctness of our behaviour. It’s just, and you know this, like anyone who turns on the television, debates here are just not comparable at all to the debates in the Chamber [of Deputies]. And not just the debates, but the focus on what’s being discussed and so on’. Senator no. 5 similarly said, ‘what I appreciate about the Senate is the higher political culture there. Because in the Chamber [of Deputies], there are all kinds of insults, attacks, and that is not the case in the Senate’.
These descriptions of the Senate, which largely correspond to reality, may influence the perceptions of ordinary citizens in different ways. Theoretically, a more sophisticated Senate environment could result in a more positive perception of the institution and, therefore, strengthen its legitimacy. Indeed, some citizens in the sample described senators as polite and respectable, among other things. However, at the same time, due to the calm and more peaceful nature of the Senate’s work, which is (at least in part) probably related to the nature of the institution itself in the political system (i.e. a reactive, corrective role, not a proactive one), the Senate is not very visible. Among other things, it does not appear often in the media, which is, in fact, typical for upper chambers (Russell, Reference Russell2001b: 452). This is described by the testimony of senator no. 14: ‘Because most of the time, this political clash is going on in the Chamber (of Deputies). And because of that clash, it is more interesting to the media. Unfortunately, nowadays, people are following more negative news, clashes etc…. Here [in the Senate], we may not agree with each other, but people here behave politely to each other’. Thus, according to both groups of respondents, the media report more on various scandals, emotions, clashes etc. In some ways, this means that Senate proceedings are more boring. As the interviews with the general public suggest, the result is that the Senate may not be very visible, and its work may not be well-known.
The second important aspect related to throughput legitimacy is the independent or even apolitical nature of the Senate in performing its roles. Senators perceive themselves as more independent than MPs, although they do not perceive themselves as non-political and do not claim that there is no influence of political parties in the upper chamber. In some interviews, nevertheless, it became clear that senators saw their relative independence as an additional source of legitimacy. This is mainly related to the aforementioned individual legitimacy deriving from personalized majority elections. A concrete, practical consequence of this kind of thinking can be found in senator no. 2’s statement that it is much harder to lobby (in a negative sense) individual and more independent senators than to influence political parties in the lower chamber.
The perception of the general public is not homogeneous in this matter. Some ordinary citizens think along the same lines, but their perception often goes even further than that of senators. Not only do they perceive the Senate as an institution more independent of political parties, but many citizens in the sample perceived the Senate as an apolitical institution. Decisions are expected to be made based on the objectivity, wisdom, experience, and independence of senators (consistent with the perceived input legitimacy described above), not from their political affiliation or support for a particular political party or ideology. Russell assumes (Reference Russell2013: 376) this view positively affects throughput legitimacy. However, according to a part of the respondents, this apoliticality and independence is only an ideal that has not been actually achieved. The fact that senators are still politicians more or less affiliated with political parties was very disappointing to some respondents, negatively affecting perceived throughput legitimacy. To illustrate this, respondent no. 20 stated:
it is not true that [the Senate] guards the republic. It’s just another segment of the political party establishment…. The Senate would be worthwhile if there were impartial people, educated experts who objectively judge things they are supposed to do, right? … But it can’t just be a group of people that’s driven by partisanship, right? That’s just nonsense.
Perceptions of throughput legitimacy can also be affected by the amount of time the institution needs for deliberation. Several senators stated that the short deadline for considering the bills referred to them (30 days) allows them to conduct rather random reviews. This is confirmed in a study by Brown and Garlick (Reference Brown and Garlick2024), who found that time and policy capacity do indeed have an impact on the quality of an upper chamber control role. On the other hand, several citizens said that the Senate unnecessarily delays the process of governance. ‘It takes ages. Instead of getting the laws passed quickly, right, it’s taking too long’, said respondent no. 16. This perception is probably unavoidable for upper chambers. Several citizens also criticized the need for compromise among multiple actors and would instead appreciate enhanced ability to act quickly and (in their eyes) related governance efficiency. However, other interviewed citizens spoke of the detached second look that is provided by the upper chamber in a positive light; in their eyes, this strengthens the throughput legitimacy of the Senate.
The Senate’s legitimacy among a part of respondents is undermined because senators often accumulate several functions and do not devote enough time to their work. But it should be noted that voters often prefer candidates they know, and senators are often active in local politics. Moreover, local politicians who are also senators can be seen as better advocates for the interests of their communities.
Output legitimacy
Output legitimacy was crucial for the general public. The legitimacy of an institution was perceived by respondents primarily in terms of the visible benefits of the institution to citizens. For a large part of the ordinary citizens in the sample, however, the benefits of the Senate were not evident. Simply put, ordinary citizens often did not know what the Senate does, why it exists in the first place, or why they should pay the salaries of senators from their taxes. In such a case, the institution can hardly be perceived legitimate. The very low level of popular knowledge of the Senate thus seems to influence the perception of the ordinary citizen.
A different perception of interviewed ordinary citizens is one that consists of a specific criticism of the Senate’s output. According to these respondents, the Senate is redundant in the process of governance because in their eyes it does not have important powers (or does not use them), and its decisions can be overridden by the lower chamber. This perception runs counter to the view that the Senate inhibits the governance process. In the interviews, both of these conflicting narratives were identified.
On the contrary, from the Senators’ point of view, the fact that the lower chamber accepts a relatively large number of their proposals for amendments confirms their existing output legitimacy. However, according to several senators, the general public is often unaware of this, which corresponds to the citizens’ frequently expressed perceptions. Even more important for output legitimacy perceived by the senators is that the Senate actually acts as a stabilizer in the system and successfully protects the democratic order of the country. According to several senators in the sample, the existence of the Senate is one reason why democratic backsliding has not occurred in the Czech Republic, unlike many other ECE countries.
However, this merit of the Senate is not very visible to most of the general public (as some senators are aware). This may be one reason why the general public does not appreciate the Senate’s role in guarding constitutionalism. For only a few ordinary citizens in the sample (who demonstrated above-average political knowledge) was the output legitimacy of the Senate strengthened because of its role in guarding the democratic constitutional order. In this respect, the perceived output legitimacy of interviewed senators and citizens differs considerably.
Nevertheless, according to some members of the general public in the sample, the Senate’s output legitimacy is strengthened by the fact that, in their eyes, the Senate actually controls the Chamber of Deputies (and potentially other politicians), and can ‘put the brakes on the MPs when things [policies, not regime in general] are going in the wrong direction’ (respondent no. 27), or ‘block against stupidity in the legislation’ (respondent no. 26). This perception is largely related to the perceived apolitical nature of the upper chamber, with many respondents likening the Senate’s potential outcomes to the decisions of an independent judicial institution. However, citizens were usually unable to name specific outputs corresponding to their narrative. In this regard, the perceptions of ordinary citizens are evidently based on imagination rather than firm opinions backed by empirical evidence.
Personal political preferences can also influence citizens’ perceptions of the Senate’s output. If the Senate limits the actions of a government with which the citizen identifies, the Senate’s output will understandably be perceived more negatively. Conversely, if the citizen is critical of the governing coalition, the outputs of an oppositional Senate will be perceived more positively. However, due to the aforementioned ignorance of the Senate and the continuous electoral cycle, some ordinary citizens do not know which political forces control the Senate at a given moment.
For some senators in the sample, output legitimacy is also apparently influenced by the work they do for their constituency. This means that they fulfil the role of a local ombudsman with their outputs. However, this relates primarily to the legitimacy of individual senators, not the legitimacy of the Senate as a whole. Moreover, the vast majority of the interviewed citizens did not know who their senator was.
Conclusion and discussion
This research clearly shows that the perceived legitimacy of an upper chamber can be affected by multiple factors and should not be reduced to democratic legitimacy gained through direct elections, which is a limitation of some previous studies (Lijphart, Reference Lijphart2012; Vercesi, Reference Vercesi2017). In this respect, I thus share Farrington’s (Reference Farrington2015) critique of McLean (Reference McLean2010) in the dispute over (non)electoral legitimacy in the context of reforming the UK House of Lords. While McLean is a proponent of the dominant narrative that legitimacy derives only from elections, this study supports Farrington’s (Reference Farrington2015: 298) point by demonstrating that an unelected chamber of parliament, in fact, may possess perceived legitimacy and can play an important and legitimate role within a wider system of government. Thus, if the UK wants to strengthen the legitimacy of the House of Lords, the introduction of direct election is not the only (or even guaranteed) solution.
To be specific, this case study clearly demonstrates that senators’ connection to their constituency and their advocacy of its interests – as well as specific qualities such as wisdom, experience, or independence – can serve as powerful sources of input legitimacy. This suggests, among other things, that upper chambers based on territorial representation may hold greater potential for being perceived as legitimate. These sources of legitimacy may balance or even overshadow the importance of (missing) democratic legitimacy. This study thus confirms that the different components of legitimacy may to a large extent compensate each other, as Hibbing and Theiss-Morse (Reference Hibbing and Theiss-Morse1995) described using the example of procedural legitimacy. In addition, such compensation may occur not only between the three main components of legitimacy, but also within each component, as illustrated for input legitimacy above.
Interestingly, interviewed senators did not perceive low turnout as undermining the legitimacy of their institution. Their perception is contrary to the arguments of many authors (e.g. Blondel et al., Reference Blondel, Sinnott and Svensson1998; Franklin, Reference Franklin1999; Lutz and Marsh, Reference Lutz and Marsh2007; Salisbury, Reference Salisbury1975), who have traditionally claimed that low voter turnout causes low legitimacy. Instead, the senators argued along the lines of Saunders (Reference Saunders2012), who considers elections to be more or less democratic depending on whether citizens have the opportunity to exercise power, and who believes that the ‘right to vote can be valuable, even if it is not actually exercised’. This thus illustrates the limited importance of general democratic legitimacy from the perspective of an institution’s members.
Even though the personal characteristics of senators and the less partisan nature of Senate proceedings both have a positive effect on the general public’s perception – as suggested by Blom (Reference Blom, Hans, Blockmans and Schepper1992) and Vibert (Reference Vibert2007) in the former case, and by Russell (Reference Russell2013) in the latter case – the crucial influence on perceived legitimacy from the general public’s perspective seems to be the output of the institution. Simply put, if an institution is to be perceived as legitimate by ordinary citizens, it should somehow benefit them and society in general. This is in line with the results of Christensen et al. (Reference Christensen, Himmelroos and Setälä2020), who, using an experimental design in a case study of Finland, showed that although the involvement of experts and citizens in policy decision-making can boost perceived legitimacy among citizens, outcome favourability is of primary importance.
However, in the case of less visible upper chambers, it may be challenging for ordinary citizens to know the actual outcomes and benefits of upper chambers due to their shallow level of knowledge of these usually controlling and rather passive institutions (Russell, Reference Russell2001b: 452; Hruška, Reference Hruška2023). The same conclusion was supported by a representative survey on the Czech Senate, according to which its role as a democracy safeguard was perceived positively, but most citizens did not know how such a democracy safeguard works (Pancíř and Čihák, Reference Pancíř and Čihák2020). Nevertheless, the way to increase the legitimacy of upper chambers from the perspective of citizens could be achieved through greater awareness-raising about the institution.
This paper also contributes to the literature on the elite–public gap in political behaviour. It demonstrates the existence of a significant gap in the perceived legitimacy of the upper chamber between senators and ordinary citizens. This finding stands in contrast to Kertzer’s meta-analysis (Reference Kertzer2022), which tends to downplay the relevance of elite–public divides. A key explanation for the gap observed in this study appears to lie in differences in knowledge about the institution between the senators and the public. That such a gap in perception may be rooted in disparities in knowledge has already been demonstrated by Hafner-Burton et al. (Reference Hafner-Burton, Hughes and Victor2013) and Riker (Reference Riker1995) in other contexts. My study therefore contributes additional empirical support to this line of inquiry. Moreover, the analysis clearly showed that the perceptions of senators, as a narrower and more knowledgeable elite group, are much more homogeneous compared to those of the general public.
Given the nature of upper chambers, which tend to be less visible and less well-known (Russell, Reference Russell2001b), differences in institutional knowledge between members of the institutions and the public may represent a broader pattern that contributes to gaps in perceived legitimacy across upper chambers more generally. The observed gap in the perception of bicameralism also provides a strong argument for future research to focus more explicitly on whose perception of legitimacy matters most. I argue that it is the perceived legitimacy among senators themselves that is particularly important for the nature of bicameralism, as they are ultimately the ones who decide whether and how to act.
As was demonstrated in this paper or as Russell showed in the case of Canadian bicameralism (2013: 381–83), perceptions can change over time, even without legal reforms. Dawson and Hanley (Reference Dawson and Hanley2019) stress that foregrounds elites’ understanding discourse about the institutions they work in is better attuned to the politics of institutional change. This is in line with the argumentation of discursive institutionalism (Schmidt, Reference Schmidt2008: 314), according to which political institutions and the actors operating within them can be understood as mutually constitutive. This is not to say that the general public’s perception is not important. However, future research should definitely consider the self-perception of upper chamber members – a topic that has been largely neglected. It is particularly important since upper chambers may be very well suited to our present populist times (Russell, Reference Russell, Albert, Baraggia and Fasone2019), and those who occupy them could provide (if they feel legitimate to do so) an important source of democratic resilience.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1755773925100131.
Data availability statement
Data (anonymous interviews) available on request due to privacy/ethical restrictions.
Acknowledgments
This research was written at the Department of Political Science, Faculty of Social Studies, Masaryk University, Joštova 10, Brno, and was supported by the Operational Programme Research, Development and Education (OP VVV) project of the ‘Internal Grant Agency of Masaryk University’ reg. no. CZ.02.2.69/0.0/0.0/19_073/0016943, and by a Specific University Research Grant provided by the Ministry of Education, Youth and Sports of the Czech Republic.
Funding statement
I would like to thank Professor Stanislav Balík for taking on the role of second coder.
Competing interests
The author declares no conflicts of interest.
Ethical statement
Respondents signed an informed consent before the interviews.
The Masaryk University Ethics Committee decided that the study would require no ethical approval.
Open access
