Shared Rule vs Self-Rule? Bicameralism, Power-Sharing and the ‘Joint Decision Trap’

Abstract

In federal and regionalised states, bicameralism constitutes shared rule between levels of governments. At the same time, second chambers serve as a safeguard protecting selfrule of decentralised governments against the encroachments of central legislation into their areas of responsibility. Both functions seem to be best fulfilled in legislative systems requiring joint decisions of legislative chambers. Depending on particular conditions, joint decision-making involves the risk that legislation ends with ineffective compromises or even fails. Under favourable conditions, it provides a productive structure to apply shared rule and protect self-rule. Comparative studies can identify these conditions, and appropriate ways to adjust institutional designs of bicameralism accordingly, bearing in mind that significant institutional reforms of bicameral systems are difficult to achieve.

Full text

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 30
ISSN: 2036-5438
Shared Rule vs Self-Rule? Bicameralism,
Power-Sharing and the ‘Joint Decision Trap’
by
Arthur Benz
Perspectives on Federalism, Vol. 10, issue 2, 2018

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 31
Abstract
In federal and regionalised states, bicameralism constitutes shared rule between levels
of governments. At the same time, second chambers serve as a safeguard protecting self-
rule of decentralised governments against the encroachments of central legislation into
their areas of responsibility. Both functions seem to be best fulfilled in legislative systems
requiring joint decisions of legislative chambers. Depending on particular conditions, joint
decision-making involves the risk that legislation ends with ineffective compromises or
even fails. Under favourable conditions, it provides a productive structure to apply shared
rule and protect self-rule. Comparative studies can identify these conditions, and
appropriate ways to adjust institutional designs of bicameralism accordingly, bearing in
mind that significant institutional reforms of bicameral systems are difficult to achieve.
Key-words
bicameralism, federalism, joint decision-making

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 32
1. Protecting self-rule by shared rule – Functions of bicameral systems
Federalism has been defined as a polity combining self-rule of central and regional
governments and shared rule among these governments (Elazar 1987: 5). This definition
also applies to federalising states in which regions, or some regions, have achieved a
significant extent of autonomy to govern their affairs. Self-rule rests on a separation of
powers between levels of government defined by the constitution. Shared rule usually finds
expression in intergovernmental relations, which exist in a variety of institutionalised or
informal patterns.
Regarding these two dimensions of a vertical division of powers in federations of
regionalised states, a bicameral system can fulfil two functions: First, second chambers
representing regional interests should protect the autonomy of regional governments
against encroachments by central legislation into their constitutionally guaranteed
jurisdiction. Thus, they serve as a safeguard of self-rule (Bednar 2009). Second,
bicameralism allows representatives from constituent units of a federation or from regions
to participate in central legislation in order to ensure that particular interests of these units
are considered. In this way, bicameral legislatives work as a core institution of shared rule.
Whereas self-rule can be effectively protected by a second chamber’s veto against
legislation, shared rule requires the willingness of both chambers to find an agreement.
Regardless of which function prevails, bicameralism essentially aims at limiting the power
of the centre.
These two functions are only fulfilled in an effective way in federations where the
legislative chambers actually represent both the demos and the demoï of constituent units
or their governments and where they are linked by processes of joint decision-making.
Whether the first condition applies depends on the selection of representatives, their
affiliation to parties, communities of citizens or governments, and the extent to which
these affiliations influence debates and decision-making in a chamber. As a rule,
I
the first
chamber consists of directly elected representatives of the demos, which includes citizens
on an equal basis. In federations, members of second chambers are normally selected on a
territorial basis and are expected to represent sub-federal units, i.e. regions, communities,
states, Länder or provinces (Borthwick 2001; Russell 2001; Palermo 2018).

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 33
The second condition depends on the decision rules in a bicameral legislature. Effective
protection of self-rule and shared rule requires that both chambers participate in legislation
on an equal footing so that they have to harmonise their decisions, which actually means
that they make joint decisions. Joint decision-making, which in bicameral system is
attributable to the veto power of a second chamber, strongly limits central power and
constraints the power of a majority in parliament. As they share legislative powers, two
chambers have to vote in favour of a bill in order to pass it as law, otherwise, legislation
fails, and neither the federal nor the regional governments can make law, the former
because of a legislative deadlock, the latter because it has no power to legislate. In order to
come to concordant votes in both chambers, a bill has to be negotiated between leaders of
majority parties or groups in both chambers, often including members or groups who are
pivotal for achieving the required votes.
In legislatures with second chambers having no veto rights, these latter can nonetheless
compel the first chamber to a joint decision because of an overlap of tasks fulfilled by
central or regional governments. This interdependence of policies, despite a separation of
powers and the self-rule of governments, regularly appears in the executive, in particular
when governments provide public goods or services, which generate cross-border
(external) effects. In legislative decisions, both levels of governments are actually affected
in matters of institutional reform or if the allocation of powers is at stake. In these policies,
most second chambers with consultative rights in normal legislation have veto rights as far
as a constitutional amendment is necessary. Hence, bicameralism usually institutionalises
‘intragovernmental’ joint decision-making at the federal level, whereas intergovernmental
relations in the executive mostly arise to manage interdependence in an institutional setting
of self-rule with shared-rule resulting from informal coordination.
In brief, bicameralism protects self-rule through shared rule; thus, it conforms to a
basic principle of a constitutional government. It prevents a concentration of power and
protects interests of minorities against what Alexis de Tocqueville has called a ‘tyranny of
the majority’ (Tocqueville 1835: 410). Power limitation is most effective if both chambers
have to come to joint decisions. Under these conditions, a second chamber can effectively
protect the autonomy of regional governments by using its veto power against legislation
threatening regional self-rule. Preventing a bill from becoming law can also help minorities
to defend their particular interests against general regulation. Failure to pass a law on a

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 34
matter, which is on the legislative agenda, may be also be an outcome preferred by
opposition parties in parliament. However, non-decisions undermine the effectiveness of a
government and can destabilise a political system. If limitation of governmental power goes
too far, legitimacy can suffer. A way out of this dilemma is negotiated legislation, which is
regularly preferred to a deadlock, even if a compromise often is considered as a suboptimal
outcome. For this reason, a bicameral system needs to optimise the limitation of power, it
also must work in a way that prevents shared rule from ending in the joint decision trap.
In order to shed light on this challenge and to make recommendations for coping with
them in practical politics, we need to understand how joint decision-making in bicameral
legislatures works, how policies are made in the chambers and how both come to a joint
decision. The theory of joint decision-making (Scharpf 1976; Scharpf 1988) has often been
reduced to an explanation of the trap. However, on closer inspection, we find that it points
out different modes of negotiation and various conditions affecting which negotiation
mode applies, and the particular consequences on policy-making. In the following sections,
I will outline the basic analytical categories and causal mechanisms, which might be useful
to explain the operation of bicameral systems in federal and regionalised states. Moreover,
I will explain why bicameral systems operating according to the joint decision mode are
difficult to change, and why they tend to be locked in the joint decision-trap of
constitutional policy.
2. Institutions and decision rules: Bicameral legislation as joint
decision-making
Bicameral legislatures vary in many respects, including not only scope of legislative and
supervisory powers, but also their very institutional composition (Coakley 2014; Heller and
Branduse 2014; Leunig 2009; Patterson and Mughan 2001; Palermo and Kössler 2017: 165-
176; Russell, 2013, 46-63; Riescher et al. 2010; Uhr 2006). In respect of the division of
powers, not all bicameral legislatures establish joint-decision systems; some second
chambers for instance only play a consultative role. In other cases, the vote of a second
chamber can be overruled by the first chamber but nonetheless has a constraining effect,
like in the British Parliament where the veto of the House of Lord can cause significant
delay to the passage of legislation. We also find bicameral legislatures, with both chambers

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 35
formally having equal decision rights, but with a second chamber rarely applying its veto
power in practice, as is the case in Canada where the Senate hardly ever votes against a
decision the House of Commons and takes on a more advisory role (Smith 2003).
In western democracies, the power of second chambers tends to be constrained (first
and foremost in fiscal policy) if they lack democratic legitimacy compared to the popularly
elected chamber of parliament. In bicameral systems with a directly elected second
chamber (e.g. the Senates in Australia, Brazil, Italy, Switzerland and the U.S.), both houses
of the legislature usually have equal powers. To make law, both have to pass an identical
bill. Accordingly, each chamber can veto a bill. In this case, decision rules require joint
decision-making. In asymmetrical bicameral systems, powers of the second chamber vary
depending on issues at stake. In Germany, for instance, the assent of the Bundesrat is only
necessary whenever a law affects the jurisdiction of the Länder. Otherwise, the Federal
Parliament has the final say in legislation (see Niedobitek 2018). In Austria, Belgium and
Spain, only laws amending the constitution require the assent of the second chamber.
When both chambers are required to reach an agreement in legislation, procedures for
conciliation vary among bicameral systems. Irrespective of other rules, these procedures
usually provide for a particular sequence of deliberations and resolutions of the individual
chambers. That sequence of the legislative process can be the same in all instances or may
differ even within one bicameral system depending on specific conditions (e.g. by policy
area, type of legislative act, etc.). In any case, the configuration of decision-making
procedures has an impact on power relations in a bicameral system, although the
consequences are ambiguous. On the one hand, the chamber where a bill is tabled, and
where a first decision is made before the bill is forwarded to the other chamber, has a ‘first
mover advantage’. It can constrain the discretion of the chamber that is second in turn and
can at best amend the version of a bill passed by the first chamber. On the other hand, the
chamber with the final decision might be able to block legislation if the proposal passed in
the other chamber is considered unacceptable. This other chamber has to take into account
the risk that its decision might be void if it cannot respond anymore with a revision of a
bill.
Power is one consequence of sequential procedures; the effectiveness of joint decision-
making is another. Those actors who initiate a bill or who are the first to decide on it are
motivated to anticipate potential vetoes in the other chamber in order to reduce the risk of

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 36
a bill getting voted down. They tend toward ‘auto-limitation’ (Manow and Burkhardt 2007),
i.e. to voluntarily refrain from asserting their preferences or to pre-emptively moderate
them. This strategic behaviour is possible, provided that the interests and preferences of
the members or coalitions forming a majority in a legislative chamber are known, or can be
predicted with high probability. The type of members and the cleavage structures in a
chamber constitute relevant factors that allow veto anticipation. Yet this is not always
possible. Party affiliation provides a rather good indicator of voting behaviour of members,
but it may only constitute one factor influencing their preferences. Moreover, since the
number of parties represented in a chamber is currently increasing in many western
democracies and, as in a second chamber a majority is usually not determined by coalition
agreements but varies from issue to issue, the capacity to reliably anticipate vetoes is limited
from the outset.
Another way to find a bicameral agreement, also requiring a sequential procedure, is
mutual adjustment. If the two chambers have sufficient opportunities to respond to the
decision taken by the other chamber, as provided for by the ‘navette’ procedures, they
might be able to come to identical decisions through mutual adjustment. Yet this process
can be time consuming, which is why the number of decisions taken in each chamber is
usually limited. Another rule to avoid an endless iteration of debates and decisions applies
in France, where the prime minister can request a final vote of the parliament (Money and
Tsebelis 1997: 1994).
Irrespective of these decision procedures, legislative chambers usually coordinate their
votes in negotiations. As a rule, agreements are negotiated informally among leaders of the
majority groups, the head of government or the responsible minister and members of a
second chamber, with these negotiations often concentrating on actors holding a pivotal
position to form a majority. If informal negotiations fail, they might continue in joint
committees set up to find a compromise between divergent bills proposed by the
chambers.
More often than not, the federal executive initiates and leads informal negotiations,
since the executive drafts most bills in order to achieve the governmental program or
agenda. These negotiations are most intensive if the second chamber assembles
representatives from regional governments, like in the German Bundesrat. Here, legislation
requiring the assent of the second chamber starts with pre-legislative intergovernmental

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 37
negotiations, first among Länder governments who search for a common position (Hegele
2017) and second in different arenas (parties, conferences of ministers) where federal and
Länder governments meet and take the opportunity to avoid a veto (Schmedes 2017). In
other federations, members of committees or party leaders negotiate a compromise. A
unique constellation can be observed in Belgium. Here the Senate lost its power of veto in
2012, except in constitutional amendments, but interests of the Regions and Communities
are represented by a plurality of regional parties in the first chamber. In consequence, joint
decision-making occurs in the form of inter-party negotiations, and most important laws
result from accords settled by party leaders (Popelier and Lemmens 2015: 120-125). In the
U.S., the President tries to convince individual members of both houses of the Congress in
bilateral negotiations, while party leaders or members of committees engage in interparty
and bicameral negotiations (Owens and Loomis 2006).
These informal negotiations are often criticised as non-transparent; they prevent voters
from ascribing responsibility for decisions and as they allow politicians to shift the blame
or to claim success, they can undermine democratic accountability. However, as long as the
final decisions are made in public, informality should not be overrated since it constitutes a
decisive condition of effective negotiation democracy. Agreements can hardly be achieved
in public debates, not least since parties compete for electoral support. They require that
actors build trust and negotiate without ‘tied hands’, i.e. without being committed to
positions of a party or a government. Compromises are more likely if negotiations proceed
behind closed doors among independent actors. Indeed, for this very reason, joint
decision-making creates a dilemma between democratic accountability and effectiveness of
policy-making, but procedures in bicameral systems can establish an appropriate balance
between both exigencies.
In several bicameral legislatures, formal procedures of mediation between the two
chambers exist (Tsebelis and Money 1997: 176-208). Conference committees in the U.S.
Congress are a case in point, as is the mediation committee of Germany’s Bundestag and
Bundesrat. These committees consist of members elected in both chambers. While the
German ‘Vermittlungsausschuss’ is a standing institution with 16 members from each
chamber, the US conference committees are established ad-hoc, with members determined
by the House of Representatives and the Senate, usually selected from the responsible
standing committees (Haas 2010: 46). These committees meet to negotiate a final version

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 38
of a bill, which subsequently can only be accepted or rejected, but not amended in the
legislative chambers. Although existing as formal institutions, they also work in private in
order to shield the members from external pressure.
Informal negotiations and negotiations in formal joint committees reduce the
probability that joint decision-making ends in a deadlock. Veto anticipation and mutual
adjustment also contribute to avoiding this negative outcome of legislation, although they
imply higher risks of failure. In asymmetrical bicameral systems, these processes occur in
the shadow of a majority decision, usually by the directly elected parliament or, as the case
may be, the popular chamber. This shadow might also result from the possibility of a call
for a referendum. In France, for instance, the President can submit a bill to a popular vote,
whereas in Switzerland, it is left to the people to initiate a facultative referendum. While
decisions of parliamentary majorities are usually known, the outcome of a referendum is
uncertain. For this reason, the latter generates more incentive for members of the
legislature to come to a broad-based agreement, while in cases where the first chamber has
the final say, this incentive is moderate and depends on the power of the majority in
parliament.
Yet even without a formal requirement of joint decision-making, bicameralism fosters
negotiated legislation and consensus democracy. Second chambers without formal veto
power often acquire influence in legislation as a ‘chambre de reflexion’. They constitute an
arena where bills are scrutinised by experienced politicians who are legitimised by their
authority rather than by their affiliation to a party, government or community. Such a
deliberative chamber generally improves the quality of legislation (Smith 2003; Sturm 2015:
185-186), both by introducing what John A Macdonald, the first Canadian Prime Minister
called ‘second sober thoughts’ and by counterbalancing the confrontation among parties in
the parliament.
However, a consultative chamber cannot change the legislative process into a kind of
deliberative democracy. Regardless of the institutional setting, whether it constitutes
symmetric or asymmetric power relations between the chambers, bicameral legislation is to
a considerable extent the result of inter-cameral and intra-cameral negotiations. The quality
of these negotiations varies. To understand the operation of bicameral systems in general
and joint decision-making in these systems in particular, it is essential to consider these
variations and their causes and consequences. The following analytical categories drawn

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 39
from theories of political negotiations can guide case studies and comparative research on
bicameral legislation.
3. Confrontation, bargaining or arguing: Conditions of negotiating
agreements between chambers
Negotiations aim to find a common ground among actors who pursue divergent
interests, but are willing or compelled to harmonise their decisions through dialogue. In
general, actors taking part in negotiations wish to come to an agreement, if they voluntarily
engage in these. In joint-decision systems, they are compelled to do so. Nonetheless, they
are not forced to negotiate, but urged to find an agreement, since they are otherwise unable
to act at all. This also applies to joint decision-making in bicameral systems, although the
actors involved in legislation advocate different policies, whether for policy or vote seeking
reasons. Nonetheless, members of governments and parliaments who are accountable to
citizens presumably try to avoid deadlocks in legislation, and members of second chambers
usually prefer a compromise to rejecting a bill passed by the directly elected parliament (or
its popular chamber), even if they have the right to exercise veto power. Therefore, actors’
behaviour in bicameral legislatures is guided by mixed motives. As responsible
representatives of citizens, most of them prefer an agreement between both chambers over
voting down a bill. Nevertheless, all of them are associated to parties or groups, which
pursue different policies and want to see most of their own ideas of good regulation
becoming law.
The fact that negotiations constitute a mixed-motive game among the actors involved
makes agreements possible. Still, there is a chance of deadlock, depending on the intensity
of conflict, the behaviour of negotiating actors, the autonomy of negotiators from external
influence, their dependence on external support, or the consequences of a non-decision.
But irrespective of how these conditions materialise, negotiators will aim at compromises
or package deals. While strategies of brinkmanship are not uncommon, in most cases, these
outcomes appear better than a deadlock, not the least from the point of view of parties
holding a majority in parliament. Yet the seemingly second-best solution, compared to a
deadlock, may produce a problematic outcome. Compromises exclude all matters of
intractable dispute, with the consequence that a law finally passed in the legislative

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 40
chambers hardly improves the status quo. Package deals can come with high costs of an
agreement, and the concessions made to achieve the required votes can imply heavy
burdens for a government. This explains why empirical studies have found that joint
decision-making in bicameral systems can cause ineffective legislation (Vatter 2005; Scharpf
1988; Tsebelis 2002).
This effect does not necessarily result from institutions constraining legislation, more
often than not it appears for the very reason that actors want to escape impending
deadlock. This is the conclusion which Fritz W. Scharpf drew from his original studies on
joint decision-making in Germany and the EU (Scharpf 1988). Since then, theory has
become differentiated, inspired by comparative research (Scharpf 1997, Falkner 2011; Benz
2016b). One important conclusion is that the probability of a deadlock and of ineffective
decisions depends on specific conditions shaping the process of negotiations, i.e.
conditions which affect how actors behave and which mode of interaction prevails in
negotiations. In bicameral legislatures, two types of conditions seem to be particularly
significant: One is the impact of party politics in the second chamber, the other relates to
the cleavage structures which determine politics and voting in a bicameral legislature.
Certainly, particular events or crises can modify the negotiation behaviour of relevant
actors, but the impact of party politics and cleavage structures remains.
To explain the effect of these conditions, we can construct categories of typical modes
of negotiation as confrontation, bargaining and arguing (Benz 2016a: 33). Confrontation
occurs, if actors stick to their positions. In consequence, the probability of an agreement
decreases to the extent that these positions diverge. Bargaining evolves, if actors pursue
their interests but are willing to make concessions in order to achieve a compromise or
settle conflicts by a package deal. Arguing requires actors to give reasons for their policy
and to search for the solution of a problem or a conflict. If a solution can be justified on
generalisable grounds, it is likely to find approval among all involved participants and the
negotiations will end with a consensus.
In politics, arguing seems to be an ideal, whereas in reality bargaining and
confrontation seem to prevail. By and large, this is correct, although it is worth noting that
arguing should not be ruled out (Bächtiger et al. 2018; Elster 1998). Negotiations often
oscillate between bargaining and arguing (Holzinger 2001; Landwehr 2009), and occasional
shifts to arguing make it possible to overcome stalemate in negotiation processes.

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 41
Moreover, particular institutional conditions foster arguing behaviour. As mentioned above
second chambers, mainly consulting on bills, and including senior, experienced politicians,
tend to negotiate in this mode. Arguing behaviour is also likely to emerge in committees
where specialists, who are motivated to solve problems, meet. The same applies to private
meetings in which negotiators can discuss free from external pressure and control.
Mediating committees constitute venues where representatives tend to negotiate in the
arguing mode, and this explains why they often help to find a compromise between
politically divided legislative chambers, as mentioned above.
However, other attributes of bicameral systems can cause a confrontation of both
houses, the principal reason being party politics. Parties compete for votes and offices. The
logic of party competition requires that their policies diverge, that each party provides an
alternative to the policies of other parties, and that policies of other parties are depreciated
and debunked. Accordingly, parties interact in an antagonistic manner, and public debates
among representatives of different parties in plenary sessions of parliaments reflect this
contestation. Expression of clearly contrasting positions is an essential mode of democratic
politics, which allows voters to assign responsibilities and hold representatives to account.
However, when compelled to negotiate, actors affiliated to different parties tend towards
confrontation rather than bargaining, not to speak of arguing which seems to be unfeasible
under such conditions.
It is due to this incompatibility with negotiations (Lehmbruch 2000) that party politics
can turn joint decision-making into a futile effort. This applies in bicameral legislatures, if
members of both chambers behave mainly as party representatives and if majorities in each
chamber are controlled by opposing parties. Both conditions have shaped politics in the
U.S. Congress since the 1990s. In situations of a ‘divided government’ when the President’s
party has no majority in either the Senate or the House of Representatives or both,
legislation becomes difficult. Meanwhile the polarisation between the Democratic and
Republican Party obstructs negotiations among members of responsible committees.
Confrontation risks undermining bargaining. In the German federal legislature, like in the
U.S., opposing parties often have a majority in the houses, since in Land elections voters
tend to prefer opposition parties in the federal parliament which then have a good chance
of achieving a majority of votes in the Bundesrat. Both in Germany and the U.S., voters
seem to use mid-term or Land elections to limit the predominance of majority parties at the

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 42
federal level. The likelihood of such a constellation of a party-political divide in the
legislature can apparently be reduced if both chambers in total are elected simultaneously,
as it is the case in Switzerland (except one representative of Appenzell Innerrhoden) and
Italy, whereas different voting systems used to elect legislative chambers can cause
incongruent majorities.
However, while party politics determines policy-making in directly elected parliaments,
this is not always a decisive factor in second chambers. Apparently, party politics
predominates in directly elected second chambers, as is exemplified by the U.S., Australian,
and Italian Senates (Breton 2014; Owens and Loomis 2006; Pasquino 2002). In
Switzerland, direct democracy moderates the role of parties in the Council of the States
(Ständerat). The British House of Lords, like the Canadian Senate, represent non-elected,
consultative chambers in which party competition has limited impact, although even in the
British upper chamber, party affiliation is increasingly reflected in voting behaviour of the
Lords (Russell 2013: 94-124). In the French Senate, members form party coalitions, but
party politics does not predominate either (Ruß 2010). In federations, second chambers
generally should represent the interests of constituent units, regional communities or
territories (Russell 2001). This does not mean that territorial interests prevail, yet they may
affect behaviour of representatives although they are committed to political parties. The
German Bundesrat, often considered a party-political body, provides an interesting
example for the interplay of party and territorial cleavages. Delegates of Land governments
are accountable to majority parties in Land parliaments but they also pursue interests of
their Land, which they represent in the Bundesrat (Leunig and Träger 2012). In a similar
vein, French Senators tend to deviate from the policy of their party if it is necessary to
defend interests of local governments in legislation affecting the territorial structure of the
state (Le Lidec 2012). Hence, beyond party politics, territorial conflicts or conflicts between
distinct societies of regions shape politics in bicameral systems (Sturm 2015).
These different cleavages can reinforce each other, for instance, if the party system
reflects the divide of a multinational federation or if economic disparities or fiscal policy
conflicts ignite regional nationalism. If they persist for some time, such congruent
cleavages find expression in a disintegration of the party system, as can be observed in
Canada and in Belgium. Congruent cleavages tend to ignite confrontation, but not
necessarily between legislative chambers. In Belgium, territorial conflicts shape politics in

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 43
the federal parliament, while in Canada they are expressed and managed in
intergovernmental relations. The Spanish legislature on the other hand seems to represent a
bicameral system burdened by mutually reinforcing territorial and party-political cleavages.
However, different patterns of conflict can also combine to crosscutting cleavages.
They exist in particular in federations with integrated party systems such as Austria and
Germany, where representatives of different territories may be aligned to the same party
and where members from opposing parties may pursue the same territorial interests. The
overlap of converging and diverging interests following from crosscutting cleavages
prevents actors from maintaining their positions in negotiations and motivates them to find
agreements. Under these conditions, bargaining behaviour is most likely to prevail in
processes aiming at a coordination of decisions in bicameral legislatures, but arguing
behaviour can also occur. Bargaining makes a deadlock unlikely, but it often ends with
ineffective compromises or inefficient package deals. Crosscutting cleavages caused by
two- or multidimensional patterns of conflict make package deals more complicated since
the number of issues to be considered increases. Therefore, negotiations mostly result in
compromises unless institutional conditions favour arguing processes.
To conclude: Bicameral systems established to constrain the power of central
government and to protect self-rule of constituent units in principle fulfil their functions if
they require joint decision-making in federal legislation. Given favourable conditions, joint
decisions can be achieved in formal or informal negotiations: if the shadow of a majority
decision drives the key actors in the chambers to find an agreement; if committees for
mediating conflicts between chambers exist; or if conflicts predominating politics in each
chamber combine to crosscutting cleavages. If party politics reinforces cleavages dividing a
bicameral legislature, the need to come to joint decisions complicates legislation, reduces
effectiveness of governance and can undermine the legitimacy of a polity. Under these
adverse conditions, balancing self-rule and shared rule will most likely fail. They increase
the probability of a deadlock in federal legislation; this does not imply that power migrates
to lower level governments but that the governance of the federation is at risk. In this case,
the joint decision trap in policy-making calls for a reform of the bicameral system. Yet,
considering this special institution, a policy of constitutional reform faces even higher
barriers (Russell and Sandford 2002).

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 44
4. The joint decision trap: Can bicameral systems be reformed?
Ineffective governance, legitimacy deficit and an imbalance of power in a federation
give reasons for institutional reform. Changes in the composition of second chambers or
of decision rules in bicameral systems require amendments of the constitution. As a rule,
laws amending the constitution have to be passed in both chambers (Kemmerzell and
Petersohn 2012). In other words, reforms determined to change the institutional conditions
of joint decision-making or to eliminate the need for joint decisions in legislation are by
themselves matters of joint decision-making. In consequence, members of second
chambers can veto a reform which reduces their power or affects their other interests. For
this reason, significant changes of structures or decision rules of a bicameral system are
unlikely, although they are not impossible. Anyway, bicameralism can lead a government to
be caught in the joint decision-trap, i.e. significantly constrained by veto players and at the
same time unable to alter institutions causing these constraints (Scharpf 1988: 267-271).
There are certainly cases of reform in federations or regionalised states which passed
bicameral legislation (Benz 2016a). In some states, constitutional amendments abolished
second chambers altogether, in New Zealand in 1950, in Denmark in 1956, and in Sweden
in 1971. In the UK, the Labour government, when returned to power in 1997, made first
strides to renovate the House of Lords, although the envisaged reform remained an
unfinished project. A recent reform of the Italian Senate was rejected in a referendum, after
it had been approved in both chambers of parliament. In Ireland, both houses of the
parliament had decided to abolish the second chamber, but as in Italy, the constitutional
amendment failed in the referendum. And more examples of significant changes in
bicameralism could be added. However, there are also cases demonstrating the difficulties
of such a reform. In Canada, various attempts to turn the unpopular Senate into an elected
chamber or to limit the Senators’ term of office have failed. In Germany, a constitutional
amendment reduced the Bundesrat’s veto powers in quantitative terms, but in legislation
affecting the jurisdiction of the Länder, they still exist; however, farther reaching reform
proposals never made it on to the agenda of federalism reform. In Romania, a consultative
referendum recommended abolishing the second chamber, but the political elite was not
willing to implement the reform.

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 45
These examples suggest the hypothesis that significant reforms of bicameral systems,
including their replacement by a unicameral legislature, are more likely in unitary states, in
particular if second chambers appear as historical relics. In contrast, second chambers
appear to be enduring institutions in federal systems. Here, they are justified on normative
grounds; apart from protecting the self-rule of lower level governments, they should
represent territorial interests in shared rule at the federal level. It is evident that, in reality,
they do not always appropriately fulfil these functions (Russell 2001, 113-114), and this
article provides theoretical reasons why this might be the case. Nonetheless, proponents of
bicameralism have stronger arguments to defend them than those who make the case for
unicameralism. This does not rule out initiatives to amend a federal bicameral system. Yet,
whenever change occurs, it turns out to be moderate and hardly deviating from a path-
dependent institutional evolution.
Path-dependency points to an explanation provided by historical institutionalism. This
theory does not rule out the occurrence of change, but it is said to take place under
exceptional conditions opening a critical juncture. Yet, considering bicameralism, Kathleen
Thelen and Sebastian Karcher (2013) have revealed, in an instructive case study on the
evolution of the German Bundesrat, that critical junctures have led to continuity while the
institution changed during periods of historical evolution. The findings of their case study
are in line with the theory of joint decision-making outlined in this article. Actors in federal
bicameral systems can incrementally adjust their practice and may also agree on moderate
changes of decision rules, but they would hardly approve a far-reaching change of
structures. However, for the same reasons that explain why joint decision-making in
secondary legislation constrains governance but does not prevent decisions, institutions can
evolve in an incremental way, although they are caught in the joint decision trap of
constitutional policy. And, in the same way that the outcome of secondary legislation in
bicameral systems varies according to conditions, the success or failure of constitutional
reforms depends on specific conditions, some of which are given whereas others can be
shaped by governments or parliaments (Benz 2016a).

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 46
5. Conclusions
Bicameral systems, even those in federations, vary. In general, they establish institutions
operating according to the logic of joint decision-making. Under unfavourable conditions,
legislation can fail, ending either in a deadlock or with ineffective compromises or
inefficient package deals. However, powers of both chambers may be equal or different
(symmetric or asymmetric bicameralism) and the selection of members of second chambers
may lead to congruent or incongruent majorities in both chambers. Therefore, the variety
of bicameral systems should make scholars cautious about generalising conclusions
regarding their operation or their effects on policy-making. Not all of them require joint
decision-making in all legislative matters, and even in symmetric bicameral legislatures
where joint decisions are the rule, their impact on legislation and on the federal balance of
power depends on particular conditions. Joint decision-making is a relevant concept to
understand legislation in federal systems. Here, this pattern of negotiated legislation
appears as an effective safeguard to protect self-rule and to include regional government in
shared rule. However, bicameral systems may cause a federation or a legislature to fall in
the joint decision-trap, and they regularly prevent constitutional amendments that seek to
significantly change power relations between chambers.
This conclusion eschews providing reasons for speaking for or against bicameralism. It
implies that the varieties of institutions, actor constellations, processes and conditions need
to be taken into account. In general, bicameralism constitutes a dilemma between
constraining power and enabling policy-making by applying power. Yet in democratic
government under the rule of law, politics is always about coping with such dilemmas.
 Arthur Benz is professor of political science at the Technische Universität Darmstadt, Germany. German
politics and comparative federalism are at the centre his research. His publications have contributed to
empirical and theoretical research on effectiveness and legitimacy in national and European multilevel
governance.
I
Although directly elected parliaments should represent citizens on an equal basis, regional communities
might find a better representation in the first than in the second chamber. This holds true in parliaments with
significant regional parties, whereas the second chamber does not represent constituent units. Examples are
Belgium, Canada and the UK.

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 47
References
 Bächtiger André, Dryzek John S., Mannsbridge Jane J. and Warren Mark (eds), 2018, The Oxford
Handbook of Deliberative Democracy, Oxford University Press, Oxford.
 Bednar Jenna, 2009, A Robust Federation: Principles of Design, Cambridge University Press, Cambridge.
 Benz Arthur, 2016a, Constitutional Policy in Multilevel Government. The Art of Keeping the Balance, Oxford
University Press, Oxford.
 Benz, Arthur, 2016b, ‘Politikverflechtung und Dynamik des Föderalismus’, in Benz Arthur, Detemple
Jessica, and Heinz Dominic, Varianten und Dynamiken der Politikverflechtung, Nomos, Baden-Baden.
 Borthwick R.L., 2001, ‘Methods of Composition of Second Chambers’, The Journal of Legislative Studies,
VII(1): 19-26.
 Brenton Scott, 2015, ‘State-based Representation and National Policymaking: The Evolution of the
Australian Senate and the Federation’, The Journal of Legislative Studies, XXI(2): 270-280.
 Coakley John, 2014, ‘The Strange Revival of Bicameralism’, The Journal of Legislative Studies, XX(4): 542-
572.
 Elazar Daniel J., 1987, Exploring Federalism, University of Alabama Press, Tuscaloosa/AL.
 Elster Jon., 1998, ‘Deliberation and Constitution Making’, in Elster Jon (ed), Deliberative Democracy,
Cambridge University Press, Cambridge, 97-122.
 Falkner Gerda (ed), 2011, The EU's Decision Traps. Comparing Policies, Oxford University Press, Oxford.
 Haas Christoph M., 2010, ‘Zweite Kammer erster Klasse: Der US-Senat’, in Riescher Gisela, Ruß
Sabine and Haas Christoph M. (eds), Zweite Kammern, München, Oldenbourg (2nd edn.), 25-60.
 Hegele Yvonne, 2017, ‘Multidimensional Interests in Horizontal Intergovernmental Coordination:
The Case of the German “Bundesrat”’, Publius. The Journal of Federalism, online first
https://doi.org/10.1093/publius/pjx052.
 Heller William B. and Branduse Diana M., 2014, ‘The Politics of Bicameralism’, in Shane Martin,
Saalfeld Thomas and Strøhm Kare (eds), The Oxford Handbook of Legislative Studies, Oxford University Press,
Oxford, 332-351.
 Holzinger Katharina, 2001, ‘Verhandeln statt Argumentieren oder Verhandeln durch Argumentieren?
Eine empirische Analyse auf der Basis der Sprechakttheorie’, Politische Vierteljahresschrift, XLII(3): 414-446.
 Kemmerzell Jörg and Petersohn Bettina, 2012, ‘Regional Actors’ Participation in Constitutional
Reform: Opportunities, Limits and Risks’, in Benz Arthur and Knüpling Felix (eds), Changing Federal
Constitutions. Lessons from International Comparison, Barbara Budrich Publishers, Opladen, London, Toronto, 311-
319.
 Landwehr Claudia, 2009, Political Conflict and Political Preferences. Communicative Action Between Facts, Norms
and Interests, ECPR Press, Colchester.
 Lehmbruch, Gerhard, 2000, Parteienwettbewerb im Bundesstaat. Regelsysteme und Spannungslagen im
Institutionengefüge der Bundesrepublik Deutschland, Westdeutscher Verlag, Wiesbaden (3rd edn.).
 Le Lidec, Patrick, 2012, ‘Decentralisation and Territorial Reforms in France: How Constitutional
Constraints Impact Strategies for Reform’, in Benz Arthur and Knüpling Felix (eds), Changing Federal
Constitutions. Lessons from International Comparison, Barbara Budrich Publishers, Opladen, Berlin, Toronto, 249-
267.
 Leunig, Sven (ed), 2009, Handbuch Föderale Zweite Kammern, Barbara Budrich Publishers, Opladen,
Farmington Mills.
 Leunig Sven and Träger Hendrik (eds), 2012, Parteipolitik und Landesinteressen. Der deutsche Bundesrat
1949-2009, LIT-Verlag, Münster.
 Manow Philip and Burkhart Simone, 2007, ‘Legislative Self-Restraint Under Divided Government In
Germany, 1976–2002’, Legislative Studies Quarterly, XXXII(2): 167-191.
 Money Jeannette and Tsebelis George, 1995, ‘The political power of the French Senate:
Micromechanisms of bicameral negotiations’, The Journal of Legislative Studies, I(2): 192-217.
 Niedobitek Matthias, 2018, ‘The German Bundesrat and Executive Federalism’, Perspectives on Federalism,
X(2).

DOI: 10.2478/pof-2018-0015 VOLUME 10, ISSUE 2, 2018
© 2018. This work is licensed under the Creative Commons
Non Commercial-No Derivatives 3.0 License. (CC BY-NC-ND 3.0)
E - 48
 Owens John E. and Loomis Burdett A., 2006, ‘Qualified exceptionalism: The US Congress in
comparative perspective’, The Journal of Legislative Studies, XII(3-4): 258-290.
 Palermo Francesco, 2018, ‘Beyond Second Chambers: Alternative Representation of Territorial
Interests and Their Reasons’, Perspectives on Federalism, X(2): 49-70
 Palermo Francesco and Kössler Karl, 2017, Comparative federalism: Constitutional Arrangements and Case
Law, Hart, Oxford and Portland.
 Pasquino Gianfranco, 2002, ‘The Italian Senate’, The Journal of Legislative Studies, VIII(3): 67-78.
 Patterson Samuel C. and Mughan Anthony, 2001, ‘Fundamentals of Institutional Design: The
Functions and Powers of Parliamentary Second Chambers’, The Journal of Legislative Studies, VII(1): 39-60.
 Popelier Patricia and Lemmens Koen, 2015, The Constitution of Belgium. A Contextual Analysis, Hart,
Oxford and Portland.
 Riescher Gisela, Ruß Sabine and Haas, Christoph M. (eds), 2010, Zweite Kammern, München, Wien,
Oldenbourg (2nd edn).
 Ruß, Sabine, 2010, ‘Der französische Senat: die Schildkröte der Republik’ in Riescher, Gisela, Ruß,
Sabine, Haas, Christoph M. (eds), Zweite Kammern, München, Wien, Oldenbourg (2nd edn.), 261-288.
 Russell Meg, 2001, ‘The Territorial Role of Second Chambers’, The Journal of Legislative Studies, VII(1):
105-118.
 Russell Meg, 2013, The Contemporary House of Lords: Westminster Bicameralism Revived, Oxford University
Press, Oxford.
 Russell Meg and Sandford Mark, 2002, ‘Why are Second Chambers so Difficult to Reform?’, The
Journal of Legislative Studies, VIII(3): 79-89.
 Scharpf Fritz. W., 1976, ‘Theorie der Politikverflechtung’, in Scharpf Fritz W., Reissert Bernd and
Schnabel Fritz, Politikverflechtung: Theorie und Empirie des kooperativen Föderalismus in der Bundesrepublik, Scriptor,
Kronberg/Ts., 13-66.
 Scharpf Fritz W., 1988, ‘The Joint Decision Trap: Lessons from German Federalism and European
Integration’, Public Administration, LXVI(3): 239-78.
 Scharpf Fritz W., 1997, Games Real Actors Play: Actor-centered Institutionalism in Policy Research, Westview
Press, Boulder/Col.
 Schmedes Hans-Jörg, 2017, ‘Regieren im semi-souveränen Parteienbundesstaat. Die administrativen
und politischen Koordinierungsstrukturen in der Praxis des deutschen Föderalismus’, Zeitschrift für
Parlamentsfragen, XLVIII(4): 899 – 921.
 Smith David E., 2003, The Canadian Senate in Bicameral Perspective, University of Toronto Press,
Toronto.
 Sturm Roland, 2015, Der deutsche Föderalismus. Grundlagen – Reform - Perspektiven, Baden-Baden.
 Thelen Kathleen and Karcher Sebastian, 2013, ‘Resilience and Change in Federal Institutions: The
Case of the German Federal Council’, in Benz Arthur and Broschek Jörg (eds), Federal Dynamics: Continuity,
Change, and the Varieties of Federalism, Oxford University Press, Oxford, 117-139.
 Tocqueville Alexis de, 2010 [1835], Democracy in America, vol. I, edited by Edoardo Nolla, Indianpolis,
Liberty Funds, Indianapolis.
 Tsebelis George, 2002, Veto Players. How Political Institutions Work, Princeton University Press,
Princeton.
 Tsebelis George and Money Jeannette, 1997, Bicameralism, Cambridge University Press, Cambridge.
 Uhr John, 2006, ‘Bicameralism’, in Rhodes R.A.W, Binder Sarah A. and Rockman Bert A. (eds), The
Oxford Handbook of Political Institutions, Oxford University Press, Oxford, 474-494.
 Vatter Adrian, 2005, ‘Bicameralism and policy performance: The effects of cameral structure in
Comparative Perspective’, The Journal of Legislative Studies, XI(2): 194-215.