There are (at least) four key values or principles of democratic governance. These are: (1) effective and responsive government, (2) stable government, (3) accountable government, and (4) representative and deliberative legislative bodies. Given the trade-offs among them, democratic polities cannot achieve all of these values equally but they are expected to attain at least a “minimum core” of each and to aim at balancing or perhaps jointly optimizing them.

This goal faces both a general problem and a more specific contemporary one. The general, and long-standing, problem is the central role and importance of political parties in modern democracies. Because political parties and their leaders compete to occupy two of the major governance institutions (the executive and legislature) and exercise public power, they can concentrate such power where the same party controls both and also disperse it where it does not, regardless of the formal or constitutional relationship between these institutions.Footnote ¹ In this way, concentration of power threatens the values of continuously accountable government (i.e., not only at elections) and a genuinely deliberative legislature and, in so doing, increases the chances of various types of “misrule.”Footnote ² On the other hand, dispersal of power risks undermining the values of effective and (sometimes also) stable government. For this reason, we cannot think of state institutions alone in analyzing or designing systems of democratic governance.

This general background problem or complexity is exacerbated by specific features of the contemporary political party systems in many democracies today. Party systems should not only be thought about in terms of numbers – a single, dominant, two-party or multiparty system – or the type of political regime in which they operate – presidential versus parliamentary partiesFootnote ³ – but also in terms of certain pathologies to which they are vulnerable. So, whether and to what extent a political party system is polarized, fragmented, or subject to hyper-partisanship also affects the difficulty of balancing and reconciling the four values. Polarization and hyper-partisanship can render both effective and accountable government, as well as deliberative legislative processes, harder to achieve because there is less, or no, overlapping middle ground. A fragmented party system makes effective and stable government less likely, as it is more difficult to obtain and sustain a governing majority.Footnote ⁴ All three features undermine the political center and the types of consensus building and accommodation that tend to be important for the optimization of all four values. They also help to create the type of alienation from democratic politics “as usual” that has fueled various types of populism over the past decade.Footnote ⁵

Both the general and the special problems can and do arise in all democratic regime types, of which there are at least six (and not only three), depending on the combination of (a) form of government and (b) political party and electoral systems. These are two party/majoritarian presidential, parliamentary, and semi-presidential systems and multiparty/PR versions of each.Footnote ⁶ In all cases, the operation of both “ordinary” party politics and the special consequences of polarized, fragmented, and/or hyper partisan party politics complicates the task of balancing the four key values of democratic governance and skew polities toward either the fragmentation or the undue concentration of political power.

A recent strand within political science and constitutional scholarship has identified “semi-parliamentarism” as a new and alternative democratic regime type and also argued for its superiority to existing ones.Footnote ⁷ The precise nature of this claimed superiority (as well as some of the institutional details) varies somewhat among its main proponents and could also benefit from being further developed but, at least implicitly or in part, the claim is that it better balances the four values.Footnote ⁸ Semi-parliamentarism is presented as a distinct variation on two-party and multiparty parliamentary regimes. It is defined by the absence of the direct election of the chief executive and the existence of two directly elected and co-equal legislative chambers in which only the first can dismiss the cabinet in a no-confidence vote, while the second has veto power over legislation that cannot be overridden by an ordinary or absolute majority of the first. Importantly, its final key feature is the incorporation of different voting systems for each of the two chambers, to try and ensure that the governing party does not control both. Whereas for pure parliamentarism in either its two-party or multiparty versions, reconciling the first two values with the second two is notoriously difficult, it is potentially achievable with semi-parliamentarism.

I agree that semi-parliamentarism is a promising regime type. In this chapter, my primary aim is to explore whether the insights of its proponents can be adapted to suggest versions of non-parliamentary democratic regimes that better reconcile and optimize the four values and address the specific challenges of political party polarization, fragmentation, and hyper-partisanship. In other words, my focus is not on the question of which regime type is superior overall but rather on how to maximize the potential benefits of the semi-parliamentary model through ambitious, but not wholesale or root and branch, design reforms in the face of current democratic challenges. Pragmatically, given the well-known “stickiness” or path dependence of forms of government,Footnote ⁹ ruling out these potential benefits to the roughly two-thirds of non-parliamentary democratic polities seems like a waste. Specifically, I will argue that semi-parliamentarism’s core feature of “symmetrical” and “incongruent”Footnote ¹⁰ bicameralism is detachable from parliamentarism and that, with suitably customized modifications and reforms, is available in presidential and semi-presidential versions that may similarly reduce the contemporary pathologies of party systems and better balance the underlying values of democratic governance than existing regimes of these types. In so doing, all three adapted forms may also address some of the causes, and resist some of the consequences, of democratic backsliding in general and authoritarian populism in particular. The secondary aim of the chapter is to consider whether the design features of these versions that involve political parties and voting systems, rather than institutional powers and relations, should be constitutionalized and, if so, which.

9.1 The General and Special Problems That Political Parties Pose to Optimizing the Four Key Values of Democratic Governance

It is widely accepted that democratic governance seeks to promote at least four key values or principles. The first is effective and responsive government. Political parties and their leaders campaign during elections not simply to occupy public office but to offer voters a meaningful choice of policies on issues that matter to them that they will seek to put into effect if elected. This capacity to bridge, translate, and aggregate voters’ policy preferences into governing and legislative agendas is perhaps the central function and justification of political parties in a democracy.Footnote ¹¹ Being elected to power and obtaining the relevant majority support legitimizes one policy agenda over another and, ceteris paribus, this is what a democratic government is expected to act on: elections have consequences. The ability of a government to effectuate the policies for which it was elected, as well as deal with ongoing and unexpected issues as they arise, is the hallmark of a functional democratic polity; the inability to do so is a sign of dysfunction.Footnote ¹² The perceptions that democratic governments have been dysfunctional and/or more responsive to the interests of various elites than ordinary voters have, of course, been one of the main factors driving populisms of left and right over the past decade.

Government stability during the course of an election cycle is a second key value. Ongoing fragility or frequent turnover undermines the kind of mid-term planning that effectiveness requires and distracts voters and politicians by elevating office over policy. It also renders polities vulnerable in the face of new and unexpected crises that may arise. Obviously, too much stability is also problematic – if not always or necessarily inconsistent with democratic governanceFootnote ¹³ – as periodic (rather than frequent) turnover is another hallmark of a functional democratic polity. This, in turn, is partly driven by the third value of continuous governmental accountability, as without periodic turnover, party-state fusionFootnote ¹⁴ and entrenchment across all institutions risks impunity and the inability to meaningfully question those in power. Putting into practice this third value is one of the key functions of democratic legislatures, although it is shared with other actors, public and private, including free and independent media outlets. The other key function of democratic legislatures is to promote the fourth value, by directly representing a broader range of voters and political positions than the executive and bringing these to bear in inclusive and collective deliberation over legislative priorities and content.

These four values exist in some tension with each other and inevitably involve some trade-offs in practice, as no single democratic polity could maximize all of them. A realistic normative goal is rather to balance or jointly optimize them in a way that ensures at least a “minimum core” of each is achieved, even if certain polities afford greater weight to some than others. Indeed, each of the three widely adopted modern forms of democratic government can be thought of as designed to achieve such a balance, albeit with different emphases resulting from the particular allocation of powers and functions between the executive and legislative branches. As referenced in the introduction, the general problem is that these modern forms of government were designed (or evolved) in ways that focused only on institutional relations and either ignored or were openly hostile to political parties.Footnote ¹⁵ But the rise and role of modern political parties changes a great deal about how these forms operate in practice and complicates their actual ability to balance the values. And this general problem has been exacerbated in recent years by the particular nature of many party systems in democracies as polarized, hyper-partisan, and/or fragmented.

Starting with the general problem, institutions are occupied by leaders and representatives of political parties rather than by individuals per se, and this means that parties can effectively merge what is intended to be separate, as well as separate what is intended to be fused. This way in which parties can function like a sort of political holding company or conglomerate was well captured by Maurice Duverger in his classic work on the subject:

So, despite the classic British separation of power between King and Parliament, or its modern version between the King’s ministers in parliament and its ordinary members, as early as Bagehot it was recognized that the “efficient secret of the English Constitution” is “the close union, the nearly complete fusion of the executive and legislature powers” in the cabinet,Footnote ¹⁷ stemming from the existential need of the government to retain the confidence of parliament. With the subsequent introduction or evolution of modern political parties, party discipline, and control over their legislators, the effectiveness and stability of Westminster-style parliamentary governments was increasingly achieved at the expense of genuine accountability to, and deliberativeness of, the legislative body. But even where executive and legislative branches are designed to be more separated and independent than in modern parliamentary systems, where and when the same political party controls both, a broadly similar concentration of power occurs with analogous effects on the third and fourth values.Footnote ¹⁸ By contrast, where and when different parties control these two branches, or there is no majority party in the legislature, then accountability of the executive to the legislature and the latter’s independence to deliberate over proposed bills are often achieved at the expense of effective governance, due to the resulting gridlock and/or fragmentation of power.

This general background problem for reconciling the values of democratic governance is exacerbated by certain specific features of contemporary party systems. As referenced above, party systems should not only be categorized by the number of parties – single, dominant, two-party, multiparty, etc. – or by the regime they operate in – presidential versus parliamentary parties – but also by whether political parties are polarized, hyper-partisan, and/or fragmented. These are, obviously, distinct but overlapping pathologies. Although parties typically occupy different spaces on the relevant policy and ideological spectrum(s), polarization refers to a situation where the major parties or blocs are close to the opposite poles and far apart in their basic platforms and orientations, leaving the center of the spectrum relatively vacant. Hyper-partisanship generally references the way that parties and their supporters interact with, and treat, each other: do they engage in “hardball,” eschew cooperation and accommodation, act as if unconstrained by practical norms of bi-/multipartisanship, treat opponents as enemies or traitors, maximize the use of power for partisan ends. Although such hyper-partisanship is more likely to occur where polarization exists, it can happen without (for example, where one party breaks away from another, where the major parties cluster around a similar space on the spectrum and need to distinguish themselves, or where parties are personality rather than policy based) and need not happen with. A fragmented party system is one in which either (a) popular support is divided among several or many parties, without any one party or coalition of parties achieving majority or clear plurality support, or (b) such division takes place within, rather than between, the major parties.

These features of many contemporary democratic party systems are making the task of achieving and reconciling the four values significantly more difficult. Polarization and hyper-partisanship can render the task of forming and maintaining a government where no single party has a majority more complex and time-consuming, undermines the accountability and deliberative functions of the legislature where one party controls both branches, and makes gridlock worse during divided government. Fragmentation undermines effective and stable government, making it more difficult to obtain and sustain a governing majority and legitimate authority. Fragmentation and splintering of political power in general, and of party systems in particular, may currently be the most challenging problem bedeviling democracies around the world and has several causes.Footnote ¹⁹ These include alienation of ordinary voters from the mainstream center-left and center-right parties that have mostly governed since the end of World War II for a mix of economic and cultural reasons, a realignment of party politics away from the traditional left-right axis based on socio-economic position and educational level, and the communication revolution that has enabled new parties, individual politicians, and even single citizens to bypass traditional parties and media outlets to reach mass audiences via social media, etc.Footnote ²⁰

9.2 How the General and Special Problems Arise in All Widely Adopted Democratic Regime Types

Although the three basic and most common forms of democratic government were, in principle, intended to achieve and balance all four values, albeit in different ways, once the operative effects of electoral and party systems are taken into account, reconciliation is more difficult. Let us briefly see how and why for each of the six major democratic regime types, looking first at the “general problem” and then superimposing the contemporary special one.

In practice, the promotion and reconciliation of all four values has been hardest to achieve in parliamentary systems. In theory, as with the other forms, this is not so. The partial fusing of executive and legislative powers bolsters the effectiveness and stability of government, while still retaining full political accountability to a representative and deliberative legislature. But factoring in the impact of the electoral and party systems substantially changes the equation and balance. In two party, Westminster-style parliamentary systems resulting primarily from the majoritarian (and usually first past the post) voting system, effective and stable government is achieved at the expense of both genuine (as distinct from formal) political accountability to the legislature and inclusive, collective deliberation of the contents of legislative proposals. As indicated in the Duverger quotation above, this is due to the power and control of the typical governing (i.e., majority) party. Because of the necessary party discipline resulting from the “sink or swim together” political logic of the single election for both the executive and legislature, this regime type standardly concentrates power in the governing party, of which the prime minister is the leader, so that it typically controls the legislature through its majority. This, in turn, means that its survival is more or less assured (unless it acts in ways that cause a rebellion among its backbench members), political accountability is mostly reduced to somewhat theatrical exchanges with the official opposition party, and “government bills” that dominate the timetable are ordinarily steamrollered through the legislative process. This “ordinary” concentration of power in a majority party and its leadership (which has on occasion been referred to as an “elective dictatorship”)Footnote ²¹ has been extended and abused by authoritarian populist regimes in parliamentary systems, such as those led by Orban and Erdogan (pre-2017), to further consolidate and entrench their power by undermining all independent institutions and sources of power.

With recent fragmentations of party systems, and the resulting greater likelihood and experience of coalition or minority governments,Footnote ²² legislatures have become somewhat more independent of government control, leading to greater political accountability, representativeness (through the greater influence of smaller parties), and deliberation. But, as reflected in the chaotic period in the United Kingdom before Brexit occurred, this was very much at the expense of effective and stable government.

In other words, the United Kingdom at this time looked more like the second type of parliamentary regime, the multiparty one resulting from having a proportional representation election system. Here, and especially where there are not two blocs formed by allied parties, the traditional difficulty of reconciling the four values is the converse of the two-party version. Without a single majority party, effective and stable government can be difficult to achieve, in some cases notoriously so, but, on the other hand, the lesser concentration of power and its greater dispersal among parties may lead to a more independent and representative legislature with more scope for holding the executive accountable (including parties withdrawing support from a coalition government) and assembling ad hoc (rather than preordained) legislative majorities on particular bills. With polarization and/or fragmentation, the risks to effective and stable (coalition or minority) government are that much greater and the probabilities of ad hoc majorities for accountability or legislative purposes are smaller.

The presidential form of government, invented out of necessity in the United States, was designed to create effective and stable government through the direct electionFootnote ²³ of a legislatively irremovable single-person executive for a fixed term of office, while a separated, independent, and more representative legislature would engage in executive oversight and have the institutional freedom to exercise its major, legislative, function in a deliberative manner. Under majoritarian, two-party presidential systems, the reality has long been “separation of parties, not powers”: either a unified government where the same party controls both branches, with a high concentration of power and significant control of legislative outcomes, or a divided government with different parties in control of the two branches and the resulting risk of legislative gridlock.Footnote ²⁴ As with two-party parliamentary systems, which they resemble,Footnote ²⁵ unified presidential governments are often effective and stable but at the price of legislative accountability and deliberativeness. Divided governments are frequently ineffective, if stable, due to legislative paralysis, although oversight of (the often-increased reliance on) presidential unilateral authority is typically robust.

With polarized, hyper-partisan parties, the concerns about overly concentrated power in a unified government tend to be even greater, as bipartisan accommodation and restraints disappear, and the gridlock resulting from divided government is that much deeper and insurmountable. In this regime type, fragmentation tends to occur within, rather than between, parties so that even unified governments may be ineffective and unresponsive, as presidents find themselves unable to fulfill their legislative agendas due to internal opposition, as prominently recently in the first terms of Presidents Trump and Biden.

Apart from the United States and the Philippines, all other countries adopting pure presidentialism employ PR for legislative elections, as here the claims of representation that this voting system maximizes appear to trump the less relevant governance benefits of majoritarian systems. And yet such benefits turn out to be highly relevant as multiparty/PR presidential systems often suffer from the absence of a presidential party, or any party, majority in the legislature resulting sometimes in ad hoc support for presidentially sponsored bills in the absence of the more continuous coalitions needed to sustain the executive in parliamentary systems but sometimes in paralysis and ineffective government.Footnote ²⁶ This is one of the well-known recipes for the “Linzian nightmare”Footnote ²⁷ of presidential coups in Latin America and elsewhere. Although such ad hoc majority-building may suggest the potential for more independent, deliberative legislative processes and presidential oversight, achievement of these values is frequently undermined by weakly institutionalized, more personality-focused presidentialist parties, as compared with at least mainstream parliamentary ones. With fragmentation in particular, these particular pathologies of multiparty presidentialism tend to increase.

As the newest widely adopted form of democratic government,Footnote ²⁸ semi-presidentialism also in theory promotes all four values.Footnote ²⁹ A directly elected president who is not politically removable by the legislature ensures stability for the full length of the fixed term, even where the prime ministerial government that is fully accountable to parliament falls and changes beforehand. There are, as it were, two paths to effective government as either of the chief executives, or better both working together, can provide it. At the same time, having two chief executives produces less concentrated power than the “executive personalism”Footnote ³⁰ of the fully presidential model and, at least vis-a-vis the president, a more separated legislature to engage in both oversight and deliberation of executive legislative proposals.

As with the other forms, however, factoring in electoral and party systems often renders achievement and reconciliation of these value a more complex and difficult task. The two party/majoritarian version of semi-presidentialism risks the most highly concentrated political power of all, where the president’s party controls the legislature, for here a president is effectively (although not formally) also the head of a parliamentary party and government. In earlier work, I have referred to this possibility as “super-presidentialism.”Footnote ³¹ In this scenario, effective and stable government comes at an even higher cost in terms of accountability between elections and legislative deliberateness than in majoritarian parliamentary systems.

Where the prime minister is from the other major party, this resembles the situation in such parliamentary systems, with the exception that here, the effective leader of the opposition is the more powerful figure of the president. The timing of presidential and legislative elections, whether or not they are simultaneous, tends to be key to the probability of these two outcomes,Footnote ³² as it is with unified or divided government under pure presidentialism. With polarization, such “cohabitation” risks further undermining the effectiveness of this regime type,Footnote ³³ and fragmentation can leave even a newly elected president with reduced legitimacy,Footnote ³⁴ as well as heading an internally divided majority party, although the constraint of sustaining the parliamentary government means that these divisions are likely to play out less than in pure presidentialism. A Senator Manchin veto may doom a presidential legislative policy or nominee but not the party’s hold on governmental office itself.

Finally, multiparty semi-presidentialism raises a risk to effective government that the two-party version rarely does; namely, fragmentation of power to the extent that neither presidential nor prime ministerial authority can be sustained. This risk, which characterized the Weimar Republic for the final half of its existence,Footnote ³⁵ in a sense combines that of both other multiparty regime types. On the other hand, where this risk does not materialize and where the party system is more parliamentary than presidential in nature, legislatures may be in a better position to fulfill their accountability and deliberative functions. Again, polarization and hyper-partisanship may undermine the bases for necessary inter-party alliances and agreements, and the contemporary fragmentation of political power makes the risk of this regime type even greater.

9.3 Semi-parliamentarism

As noted above, parliamentary systems of both two-party and multiparty versions have traditionally found it hard to reconcile the four values, with each version prioritizing two different ones in ways that risk failing to achieve the “minimum core” of the other two. This is perhaps even more pronounced in the two-party, Westminster-style version because the normal majority party required for effective and stable government combined with the sink or swim political logic of the single election typically enables it to control and dominate the legislature and the legislative agenda. The resulting loss of genuine political accountability has been a major reason that many parliamentary systems have enhanced the legal accountability of government by establishing forms of judicial review for the first time.Footnote ³⁶

In the last few years, an alternative to the standard two types of parliamentary regime has been proposed that it is claimed better reconciles and optimizes the four values.Footnote ³⁷ This alternative has been labeled “semi-parliamentarism.” It is generally based on, although a modification of, the closest real-world examples, at both the national and state levels in Australia, and its main proponents have advanced a couple of different versions or sub-types. As stated above, the four defining components of this regime type are: (1) the absence of the direct election of the chief executive, (2) two directly elected legislative chambers, (3) only the first chamber can dismiss the cabinet in a no-confidence vote, and (4) both chambers have equal legislative power and exercise of the second chamber’s veto cannot be overridden by an ordinary or absolute majority of the first.Footnote ³⁸ In addition to these purely institutional arrangements, the key to the claimed advantage of this regime type in better reconciling the values is the requirement of different voting systems for the two chambers in an attempt to ensure that the same party does not control or have a majority in both: specifically, a majoritarian voting system for the first chamber, to promote effective and stable government, but a PR voting system for the second chamber, to promote accountability and legislative deliberativeness.

I find semi-parliamentarism to be an innovative and promising democratic regime type that has the potential to offer a superior version of parliamentarism to the two existing ones. Part of its merits are the internal resources it brings to bear for addressing the contemporary challenges of polarization, hyper-partisanship, and/or fragmentation that render the balancing of values more difficult. The composition and legislative powers of the second chamber create incentives for multipartisan, ad hoc, issue-specific negotiation and accommodation among represented parties that might temper polarization. In a sense, it also deals with the contemporary problem of fragmentation of power by both channeling and celebrating it. So, the attempt is to contain the inter-party version in the first chamber through a majoritarian voting system that over-rewards the two leading parties but to increase the number of parties (and so representation) in the second chamber through PR.

Among its principal expositors, Steffen Ganghof argues that semi-parliamentarism is clearly superior to presidentialism because it provides the same benefits of separated powers between the executive and (here, the second chamber of the) legislature but without the unnecessary costs of “executive personalism.”Footnote ³⁹ Although he is less clear or categorical about this, it also appears to be superior for him to both pure parliamentarism (because of the absence of separated powers) and semi-presidentialism (because most of the costs of executive personalism survive, in terms of a directly elected, irremovable chief executive, despite the existence of a second, parliamentary one).Footnote ⁴⁰ Tarunabh Khaitan also suggests that (his “moderated parliamentarism sub-type” of) semi-parliamentarism is superior to the alternatives because it “combines the most attractive elements of each” while still yielding a stable regime.Footnote ⁴¹ But he also makes the more modest claim that moderated parliamentarism is one way to optimize what he has identified as four constitutional principles relating to political parties,Footnote ⁴² as well as to balance governmental effectiveness and accountability.

For reasons of space, my aim here is not to discuss whether or not Ganghof’s arguments for the superiority of semi-parliamentarism over all three widely adopted forms of government are compelling. For what it is worth, I believe that his conception of the separation of powers is somewhat narrow and constrained; it seems to imply, for example, that pure parliamentary systems lack this value altogether.Footnote ⁴³ I also think the contrast between presidential and (modern) parliamentary systems in terms of the “executive personalism” of the former – that is, executive power being located in a single person versus a collectivity – is overdrawn in the modern era in which prime ministers are no longer simply “first among equals” and the office has become “presidentialized” in many countries.Footnote ⁴⁴ But I do think his work is insightful and illuminating, so much so that it is worth exploring whether the principles and institutional ingenuity of semi-parliamentarism are perhaps exportable to other regime types. Similarly, in terms of Khaitan’s (more modest) claim that his version of semi-parliamentarism is one way to reconcile effectiveness and accountability, I want to ask whether there might be other ways and, in particular, whether the insights of the semi-parliamentary model that he has helped to develop are capable of being adapted for other regime types. Moreover, would some of the inherent features of semi-parliamentarism that might reduce current party pathologies be exportable to these others? Are there any independent design features that might be helpful in this regard?

9.4 Does Semi-parliamentarism Suggest There May Be Ways to Better Balance the Four Values in Non-parliamentary Regimes?

As we have seen, from the perspective of optimizing the four values, semi-presidentialism and (even more) presidentialism tend to achieve stable government for the duration of the presidential term – at least, absent the “Linzian nightmare” scenario – but, depending on political party strength and alignment, risk significant underperforming on the others.

Starting with presidentialism, a bicameral legislature with the same equal powers and separate voting systems as under semi-parliamentarism would likely, with a few additional modifications, have similar potentially beneficial effects. Effective and responsive governance would be bolstered by more or less ensuring that the president’s party has a majority in the first chamber for the full duration of the president’s term. This could be achieved by combining (a) a majoritarian voting system with (b) simultaneous executive and first chamber legislative elections and (c) making the first chamber term the same as that of the president (for example, four years). Importantly, this latter feature obviously rules out midterm legislative elections that often create divided presidential government and gridlock. To the extent that ineffective and unresponsive government is driving the current alienation from the more mainstream or centrist parties, this would help to address the problem. Given the presidential party’s likely majority in the first chamber, a presidential legislative veto – as a check on a potentially hostile, runaway, or “all-powerful” legislature – is likely unnecessary,

As with semi-parliamentarism, PR elections for the second chamber will likely enhance multipartisan, and so potentially more effective, executive oversight and accountability, as well as overall legislative representation of voters’ preferences and the deliberativeness of legislative processes. With likely no presidential or any single party majority, the prospects of both rubber stamping and continuous institutional gridlock are much reduced and the incentives for multipartisan, ad hoc, issue specific negotiations and accommodations for presidentially sponsored (and other) bills greatly increased. PR thus appears to be key to the values-optimizing goals of an “incongruent” second chamber. Other institutional permutations, such as different (and possibly staggered) terms and being elected at a different time from the simultaneous presidential and first chamber elections, seem less central than under semi-parliamentarism itself,Footnote ⁴⁵ although certainly could be considered as potential supplements.

As for addressing and reducing the special problems, a number of inherent and possible additional features of such a regime hold promise. To push candidates away from the extremes and so reduce polarization (as well as increase responsiveness and representation), presidential elections should, as they do almost everywhere, require a direct national majority of voters. As the PR voting system for the second chamber is likely to result in the existence of more than two main political parties and presidential candidates, an instant or two-round run-off system would therefore need to be employed for presidential elections. Although not of course guaranteed, this increases the chances of an anti-polar majority at the second round, as we have now seen three times in France. The method of selecting party candidates for presidential (as well as legislative) office should similarly not reward more extreme or outlier positions, as, for example, the US system of party primaries can do by effectively bypassing both more knowledgeable party insiders and less motivated/partisan, ordinary party voters.Footnote ⁴⁶

At the legislative level, this type of presidential regime again inherently counters polarization and hyper-partisanship by reducing the risk of general gridlock and creating incentives for issue-specific coalitions engendering multipartisanship and accommodation. Additionally, in terms of the majoritarian voting system for the first chamber, ranked choice voting (or a run-off system) is to be preferred to first past the post, for, as a true majoritarian rather than plurality system, it increases the prospects of less extreme positions and candidates. Again, party primaries that tend to be dominated by more motivated, partisan, and extreme party members/voters are part of the current problem and should either not be the method of selection at all or replaced by a “top four”Footnote ⁴⁷ or equivalent method, to counter this effect. Independent redistricting commissions are also important, as having competitive general elections, rather than only primaries due to partisan gerrymandering, is key to reducing polarization.

In terms of fragmentation, again in a sense PR in the second chamber is designed to channel and reap its systemic benefits as compared with some of the costs of a two-party system that we have seen. If the risk or reality is of an overly fragmented second chamber exists, the common technique of capping it by employing voting thresholds for seats can be instituted. To the extent that the current fragmentation of parties and party authority is the result of the greater independence of legislative representatives stemming from the communications revolution and the potential for individual following and fundraising it has created,Footnote ⁴⁸ this independence could be reduced in at least two ways. First, a closed list PR system for the second chamber would make individuals more dependent on the party, for their ranking on the list, and in this way replace older, intra-party seniority control systems. Second, whether through law (where possible) or internal regulation, funding and campaign finance rules for all elective offices could also render politicians more dependent on party and less on their own, individual fundraising.

With semi-presidentialism, the goal of achieving greater optimization of the four values revolves around attempting to avoid both the undue concentration that occurs when a president is effectively also the head of a parliamentary government, as party leader and the undue fragmentation where no party or stable bloc has a legislative majority. Essentially the same set of institutions, powers, voting systems, and reforms as with presidentialism just discussed would increase the probability of such an intermediate outcome. With simultaneous first chamber and executive elections, same terms of office, and a majoritarian (preferably ranked choice or two round) voting system, the result is likely to be presidential control,Footnote ⁴⁹ but the key difference from the “super-presidential” scenario is the likely absence of a presidential, or any single, party majority in the directly elected and co-equal second chamber, where PR is employed. Combined with most of the other features and reforms discussed above for presidentialism, the potential result is to support effective and responsive government but without either undue concentration of power or built-in gridlock, enhance accountability to, and the representativeness of, the legislature, to create incentives for issue specific deliberation and accommodation across parties, and thereby also lower the risk of extreme fragmentation.

9.5 The Role of Constitutional Law

Democratic regime types are constituted by the combination of form of government (institutional powers and relations) and the operative voting and party system. This is why, for example, two-party and multiparty parliamentary systems are distinct types (or sub-types), generally furthering opposite values of democratic polities. It is a traditional blind spot of constitutional scholars to focus on the first part of the combination only, without the second. But a constitution should ordain and establish a regime type and not merely a form of government and so should presumptively include the key party and voting variables of the chosen type. Not only does this render it more difficult for elected officials to deliberately change or undermine a given democratic regime, but it also protects that regime from more unintended, serendipitous, or gradual alterations in its constituent parts.

As we have seen, the versions of presidentialism and semi-presidentialism that may better balance the four values and address some of the pathologies of contemporary party systems contain a distinctive set of institutional powers and relations that should presumptively be constitutionalized, as per the modern norm. To recap, these include direct national election of the president by a majority of the popular vote; two co-equal and directly elected legislative chambers, with the first having the same term as the president and a simultaneous or nearby election day, whereas members of the second chamber have different (and possible staggered) terms and/or are elected at a different time.

The key party/voting feature that is co-constitutive of these distinct democratic regime types is the employment of different voting systems for the two chambers, a majoritarian system – and preferably ranked choice voting – for the first, and PR – presumptively the closed-list variety – for the second. It is key, in significant part, because the respective party systems, two governing party contenders in the first house and multiparty in the second, largely follow from this choice. Accordingly, this should also be constitutionalized. Otherwise, depending in part on the legislative tie-breaker rule employed,Footnote ⁵⁰ it might be possible for the first chamber to repeal an ordinary statute PR requirement for the second. Given the design requirement of single-member constituencies for the first chamber, the polarizing possibility of artificially eliminating competitive general elections through partisan gerrymandering should be minimized by also enshrining independent districting commissions in the constitution. This would likely have a knock-on effect of reducing partisan incentives for holding party primaries that select more extreme candidates, so that constitutional regulation of parties in this regard may be unnecessary.

9.6 Conclusion

This chapter represents an initial exploration of the possibilities of incorporating variations of political party and electoral systems into constitutional design for the purpose of addressing the twin contemporary democratic challenges of fragmented and unduly concentrated political power. It generalizes from the insights of proponents of semi-parliamentarism to suggest that its core promising feature of “symmetrical and incongruent” bicameralism can usefully be adapted to create equivalent versions of presidentialism and semi-presidentialism. By seeking ways to render all types of democratic government more effective and responsive, the chapter aims to counter the fragmentation and dysfunction that is helping to drive the appeal of more polar political forces, including authoritarian populisms of left and right. At the same time, it also seeks to curb the overly concentrated power that single party (or alliance) control of executive and legislative branches risks. For not only does this undermine the democratic values of accountability and legislative deliberation, but such concentrated power has been exploited and abused in recent years by various authoritarian populist regimes, both those with and without clear electoral majorities. In these ways, the aim is also to address some of the causes and consequences of such regimes. Obviously much remains to be done in terms of filling in the details.