Allen Hicken
Professor of Political Science, University of Michigan

Introduction: Three questions and three principles for reformers

The purpose of this paper is to lay out some of the promise, peril, and trade-offs involved with institutional design. The context for the memo is the institutional design process currently underway in the Bangsamoro Autonomous Region in Muslim Mindanao (BARMM). The task before those involved in the design of BARMM is an enormous one—but one common to institutional engineers across the world. How do you design political institutions that work? How do you improve on existing institutions while maintaining the things that are working? How do you balance competing ideas, interests, and priorities?

Institutional reform is no panacea. By themselves institutions cannot change human nature, they cannot reshape culture, they cannot force opponents to see eye to eye, and they have no power to imbue all public officials with a moral heart and incorruptible soul. But, they are one of the few tools available to reformers—one of the few things that can be changed in the short term. Institutional reform, to be sure, can be a clumsy and inefficient tool—it is often more like a rock than a scalpel— but it is often the only tool we have.

And we know that institutions and institutional reform can be a powerful tool. Why is that? Institutions help shape a variety of things we care about. First, they shape the incentives and preferences of political actors: voters, politicians and bureaucrats. For example, political institutions can shape the incentives of political actors to cooperate or compete. They can encourage or discourage the creation of lots of political parties. Political institutions also influence politicians’ incentives to focus on programs and policies versus patronage and pork. And institutions help determine to which groups politicians are accountable. Second, political institutions help establish the relative capabilities of political actors. They determine who has a formal seat at the policy making table, for example. They set out the limits on each actors’ formal powers—can they act unilaterally, or do they have to cooperate with others?—and define the scope of their power relative to other political actors. Finally, through their influence on incentives and capabilities political institutions shape the behavior of political actors.

Given the power of institutions to shape political outcomes, it is essential that reformers approach institutional design with the utmost care and specificity, along with a healthy dose of humility. Doing so requires reformers to consider three questions, and adhere to three principles.


Three Questions

Before plunging headlong into the process of institutional design it is essential that designers ask three questions. First, what problems are we trying to solve, or what goals are we to trying achieve? Specificity here is key—we must be specific about the kinds of changes we would like to see. It is not enough to say that we want to reform the political system, or that we want better government, or more democracy. What precisely are we trying to do? What problems, exactly, are we trying to fix? Attempting institutional reform without specifying the purposes of such reform is a recipe for failure and disappointment.

If the first question requires us to identify the symptoms we are trying to treat, the second question invites reformers to think carefully about the sources of those symptoms—to move beyond description (“it hurts here”) to diagnosis (“I’ve got an inflamed appendix.”). Namely, what do we think is the cause of the problem we would like to solve? Imagine that we identify corruption as a problem we would like to tackle via institutional reform. Before proposing institutional solutions, we must first identify what we think the drivers of corruption are (and there may be more than one). Is it low salaries for public officials? Is it a lack of monitoring and oversight? Is it poorly designed, inefficient systems? Is it a culture of impunity? Only once we’ve identified the cause(s) of the problem, can we tailor an appropriate remedy.

Once we have specified the problem we want to fix (or goal we want to achieve), and then identified the relevant sources of that problem, we can ask ourselves the third and final question: How would the proposed reform remedy that problem or help us achieve our goal? This requires designers to be clear about their expectations—in effect, to describe their theory connecting the proposed institutional change to the desired outcome. Back to our corruption example, how would raising bureaucrat salaries lower incentives for corruption? Or why might strengthening parties decrease incentives for vote buying? Often in reform discussions we see cures, remedies, and recommendations being offered without sufficient attention to these three basic questions. This is the equivalent of my doctor saying “I’m not sure exactly what is wrong with you, but take this pill, it is bound to help.”


Three Principles

With the answers to these three questions in hand, reformers can then turn to the actual task of institutional design. As they begin the difficult task of institutional engineering, I suggest three principles to keep in mind. First, institutions embody trade-offs. The choice of institutions reflects decisions about which sets of goals and values are going to be prioritized over others. People will have good faith disagreements about what the priorities should be, but even if we agree on those priorities, it is often impossible to maximize all of our goals and priorities at once. For example, let’s say my two priorities are a) more effective government (government that can act quickly and decisively in the face of a crisis), and b) government that is more responsive to public demands.

Those two goals may be in tension. Allowing the executive branch to act unilaterally in the face of a crisis may maximize decisiveness, but come at the cost of some responsiveness. Likewise, requiring governments to consult extensively with the legislature and public stakeholders before acting will improve responsiveness, but at the cost of quick action. Capable reformers recognize the tradeoffs embedded in institutions and try and strike a purposeful balance, consistent with their goals and priorities.[i]

Second, institutional reforms do not take place in a vacuum. As social scientists we often think in terms of experiments conducted on the basis of all else being kept equal. What happens if we keep everything the same, and just change one thing? This is useful for theory building and empirical analysis but in reality it almost never happens that way. Often, reforms come as a package—there are lots of things being done at once. Some of these institutions will reinforce each other. In other cases, the goals of one set of institutional reforms may be undermined by reforms in other areas. Institutional reforms work best when they are carefully calibrated to work together as a package, and when they are in balance with existing institutions, norms, and cultural attitudes.

The implication of this is that institutional reform is a difficult, messy process. Reformers can predict certain things, and have strong expectations about the effects of reforms, but there are always things that reformers do not anticipate. There will be unexpected events that occur. Actors will behave in ways that were not fully anticipated. And the interaction between the new institutions and the larger political and social context will unfold in surprising ways. In other words, there are always unintended consequences.

Third, where possible, institutional designers should keep it simple. There is a real temptation as an institutional engineer to try bold new things and to try and change a whole bunch of things at once. Sometimes that is necessary. But remember that the more comprehensive, complicated, and novel reforms are, the greater the chances that things will interact in ways that reformers did not expect, the greater the chance that we will see serious unexpected consequences, and the greater the chance that reforms will fail to achieve the goals of reformers.

This comes back to stating very clearly what our goals are. What is the problem we are trying to solve with a given reform, or the outcome we are trying to achieve? How would the proposed reform help get us to that outcome? Are there simpler ways to get the same result? Keep in mind this quote from Albert Einstein: “If you can’t explain it simply, you don’t understand it well enough.” If an advocate for a given reform cannot clearly explain what problem the reform is designed to fix and exactly how the reform would fix that problem, then we should be wary.

The remainder of this paper proceeds as follows. In the next section I review the institutions as set up in the Bangsamoro Organic Law (BOL) and then note sets of decisions still to be made that have the power to shape how these institutions function in practice. First, I explore questions surrounding the mixed-member electoral system, specifically the electoral rule for the party tier (closed or open list proportional representation), whether the tiers are linked, and how reserved seats will be filled. I note that these choices have profound effects on whether or not the mixed-member system will help promote stronger parties, as intended, and minimize the degree of party system fragmentation. Second, I discuss design and functioning of the proposed parliamentary system, the most notable innovation in the BOL, including questions about the rules governing government formation and dissolution. I also emphasize the importance of pairing parliamentarism with a supportive party system, specifically one with a modest number of strong political parties, and review institutional choices that will help shape the number and strength of BARMM’s political parties.


The Institutions of the BOL

The institutions of the BOL represent a major departure from the political institutions elsewhere in the Philippines. Understanding the national political context is essential when we approach designing institutions for BARMM. BARMM will sit within, interact with, and be shaped by this broader political-institutional environment (and perhaps, in turn, help reshape that environment). Many of the institutional reforms being proposed appear to be in response to perceived weaknesses in the broader Philippines’ political system, and thus BARMM has an opportunity to correct some of the dysfunctional characteristics of that system.

My review of the BARMM institutions will focus on a) the nature of the electoral system, and b) the choice a parliamentary system. In each case I will highlight unresolved questions that still need attention, and discuss supporting institutions needed to help these function effectively (specifically, the party system).


The Electoral System

The BOL establishes a mixed-member electoral system, with three categories of seats, and at least two separate methods of election to fill 80 total seats. Up to forty percent of the seats are to be filled using the familiar plurality (“first-past-the-post”) rule within 32 single-member districts, with electoral districts having a minimum population size of 100,000. This is called the “district representative tier.” At least fifty percent of the seats are to be filled using proportional representation from a single region-wide district with 40 seats. This is called the “party representative” tier. Finally, at least ten percent of the seats are reserved for particular groups. Specifically:

    • 2 seats for non-Moro indigenous groups
    • 2 seats for settler communities
    • 1 seat for women
    • 1 seat for youth
    • 1 seat for traditional leaders
    • 1 seat for the Ulama

A number of details about the electoral system are still to be worked out. These include:

For the Party Representative (PR) tier:

    • What electoral formula will be used?
    • Will party lists be open or closed?
    • What will the electoral threshold be, if any?

For the District Representative (plurality) tier:

    • Must candidates be affiliated with a political party in order to stand for election? In other words, are independent candidacies allowed? (Article 7, Section 20 seems to suggest members of parliament (MPs) do not need to affiliate with a )
    • The BOL sets 100,000 as a population minimum for plurality districts. What is the ceiling? In other words, how much malapportionment will be tolerated? For example, if a municipality has more than 200,000 in population (e.g. Cotabato) does the law require it to be divided into two electoral districts?

For the Reserved and Sectoral Seats:

    • What is the method for selecting the reserved and sectoral seat representatives?
    • If the seats are elected, who is eligible to cast a vote for each group?
    • If the seats are elected, who is eligible to stand for election for each group? Note, the BOL states that youth representatives must be between 18-30, but does not require representatives of the other groups to come from those

Questions for all tiers:

    • Are there any geographic requirements? Must parties run candidates in a certain percentage of districts? Must they win a certain percentage of the vote in all or some areas to win seats?
    • Will the different electoral tiers be linked in any way?
    • Can candidates stand for election in both the Party and District tiers?
    • Can parties field candidates in both tiers?
    • What counts as a “regional party” as per Article 7, Section 9?

Let me highlight the trade-offs and risks involved with a few of these questions. Specifically, the ways in which choices about the electoral system will shape the strength and number of political parties. I will then discuss alternatives for filling the reserved seats.


Closed or Open List PR and the strength of political parties

In its choice of electoral system the BOL follows the footsteps of a growing number of countries that have adopted similar systems in a bid to avoid some of the extremes and pathologies of electoral system design. This type of system is called a mixed-member system, in reference to the fact that members of parliament are elected in a variety of different ways. One of the chief advantages of a mixed-member system is that it tries to strike a balance between valuing close ties between individual politicians and local voters/constituents on the one hand, and encouraging strong parties and party loyalties (party-centeredness) on the other. The district plurality tier protects the voter-politician link, and ensures there are politicians with ties to particular geographic constituencies, while the party list proportional tier provides incentives for investing in building and using strong parties. However, this balance only works if the choice of electoral system for the party list tier promotes party-centeredness. The almost universal choice of electoral system for the party list tier—closed list proportional representation (CLPR)—does just that. CLPR puts the power of ballot access and placement in the hands of party leaders, while requiring voters to cast their vote only for the party of their choice. The result is a system where both voters and candidates are likely to be oriented around political parties and their differences, encouraging voters to prioritize party label when making voting decisions, and requiring that the narrow short-term interests of politicians/candidates are balanced with the broader, longer-term interests of the party as a whole.

By contrast using a candidate-centered electoral system, such as open list PR (OLPR), would be problematic. OLPR gives voters the option of voting for candidates on a party’s list, and the votes for candidates determine the order in which a parties’ candidates are seated. By pitting candidates from the same party against each other, OLPR undermines the importance of the party label for both candidates and voters. While OLPR has its advantages, it is also associated with a litany of problems, including higher levels of particularism and patronage (e.g. Indonesia) and higher levels of corruption (Aspinall 2019). OLPR is rarely employed in parliamentary regimes outside of Europe, where there is already a historical foundation of strong political parties. If used as part of a mixed- member electoral system, OLPR would undermine the chief advantage of such a system: providing incentives for investing in political parties. Instead, OLPR would amplify the candidate-centered nature of the district plurality elections.


The relationship between the district and party tiers and the number of parties.

Given the risk that a highly fragmented party system poses in a parliamentary system, as discussed in the next section, the drafters of the Electoral Code will want to be cognizant of the ways in which different rules could increase or lessen fragmentation. One of the biggest outstanding questions in the design of the BARMM mixed-member electoral system pertaining to the number of parties is whether or not district and party tiers will be linked in any way. Those tiers can be linked through both the electoral and party systems. Let’s start with the party system.

A key question for drafters of the electoral code is whether the tiers are linked in terms of partisanship. In almost all mixed-member systems the same parties can (or may be required to) field candidates in both the district and party list tiers. As already discussed, one purpose of a mixed system is to provide a balance within parties between local and broader interests, and between party and candidate preferences. The national Philippines electoral system is unusual in this regard in that it prohibits the mainstream parties from running for seats on the party list, and there is some discussion about doing the same in BARMM. Let me state again how unusual it is to mandate completely different party systems for each tier. This bifurcation is associated with a number of problems at the national level in the Philippines (see Hutchcroft 2019; Teehankee 2019), but here I want to focus on the effect of such an arrangement on party system fragmentation. At the national level, the party list system has contributed to the fragmentation in the party system post-Marcos, though this has been partially offset by the small size of the list tier and the party switching by other members of congress. By contrast, the party list tier in BARMM is much larger, relatively speaking, and party switching is prohibited. To be clear, restricting parties to competing in only one tier is guaranteed to balloon the number of parties in parliament. As I will discuss in more detail below, a large number of political parties carries greater risks under parliamentarism than under presidentialism. Under both presidentialism and parliamentarism party system fragmentation can cause gridlock, but under parliamentarism fragmentation can also undermine government stability. If a goal is to keep the number of parties to a modest number, conducive to government formation, cabinet stability, and policy making, parties should be allowed (or even required) to field candidates in both the district and party list elections.

Before turning to other ways the district and party tiers could be linked, let me return to the earlier discussion about promoting party-centered politics. Rules about which parties are allowed to compete in each tier are also important here. In mixed-member systems the hope is that the presence of the party list tier will encourage all parties to invest in party building. However, if parties are segregated by tier, with different parties in each, then there are weaker incentives for parties and candidates competing only in the district tier to invest in or rely on party labels, and a concern that two classes of MPs might emerge. Indeed, the problematic experience of the party list elections at the national level in the Philippines is a cautionary tale. By walling off the party list seats from regular party competition, and by capping the number of seats party list parties could obtain, the electoral system has inadvertently contributed to the further marginalization of already marginalized groups (Teehankee 2019).

If the potential demarcation of political party competition across district and party tiers is an issue quite distinctive to the Philippines, there is another issue common to mixed-member systems throughout the world. That is how the district and party tiers may (or may not) relate to each other in the actual allocation of seats. Accordingly, mixed-member systems can be divided into two broad categories. In parallel systems (also called mixed-member majoritarian systems) the district and party tiers are separate and distinct. The results from one tier don’t affect the results in the other. Japan, Korea, and Thailand (2001-2017) are examples of this kind of system. By contrast, in compensatory systems (also called mixed-member proportional systems) the results of one tier affect the distribution of seats in the other—the tiers are linked in other words. Specifically, the goal is to ensure that the final seat distribution reflects the distribution of party list tier votes. If parties are under-represented in the district tier results, the party list seats are used to compensate for that disproportionality. Germany and New Zealand each use compensatory systems. In general, a compensatory system is more favorable to small and medium-sized parties, and is thus associated with a larger number of parties.

The table below contains a comparison of how parallel and compensatory mixed-member systems work. In this example the number of seats in parliament is 70, with 30 allocated to the district tier, and 40 to the party list tier. In this election Party A wins 20 of the district seats and Party B wins 10. They split the party list votes 50:50. Under a parallel system we would divide the party list seats in half, 20 for each party, and add that to their district seat total. So, Party A ends up with 40 seats (20 district seats plus 20 from the party list) and Party B gets 30 (10 + 20). Under a compensatory system, by contrast, we would use the 40 party list seats to compensate for any disproportionality in the district tier by making sure that the final seat totals reflect the distribution of votes in the party list election. Since party A and B each got 50 percent of the party list votes, we want to add enough party list seats to each party so that they end up with 50 percent of seats in parliament. Party B would thus get 25 additional seats, and Party A 15 seats, giving them both 35 total seats.

District Seats Party List Vote Share Parallel Compensatory
Party A 20 50 40 35
Party B 10 50 30 35

The following summarizes some of the tradeoffs associated with a parallel versus compensatory mixed-member system.


    • Tends towards fewer parties by favoring larger parties
    • Less proportional
    • Simple to understand and calculate


    • Tends towards more parties by favoring small and medium-sized parties
    • More proportional
    • Harder to understand and calculate

Finally, the BOL does not specify whether the parties competing in the party list tier will be subject to a threshold. Many countries that employ proportional representation apply electoral thresholds— parties that fall below a certain percentage of the votes do not receive seats. The higher the threshold, the harder it is for smaller parties to win representation. Countries adopting such thresholds are willing to sacrifice some representation and proportionality in order to reduce the number of parties, and make it harder for extremist parties to win seats. Thresholds vary from less than 1 percent, to more than 10 percent, but 3-5 percent is the most common range. (For example, Indonesia’s electoral threshold has been raised over time and currently sits at 4 percent.)


Reserved seats

Perhaps nothing in the BOL raises more questions than the provision reserving at least 10 percent of the seats (8 seats) for underrepresented and sectoral groups. As a reminder, those reserved and sectoral representative seats are broken down as follows:

    • 2 seats for non-Moro indigenous groups
    • 2 seats for settler communities
    • 1 seat for women
    • 1 seat for youth
    • 1 seat for traditional leaders
    • 1 seat for the Ulama

How these seats are to be filled is left to the BTA (for the first election) and parliament thereafter (Article 7, Section 7).[ii] In the previous section I posed a number of questions related to reserved seats that drafters will have to address. In this section I describe different options for filling those seats, along with some pros and cons for each. Note that the most common way of filling reserved seats, namely awarding each party a number of the reserved seats proportional to their vote or seat share in the general election, is not an option for BARMM. There is only one seat for most groups, and you cannot divide a single seat proportionally.

  1. Seats are filled by the cabinet or parliament: After the election is complete, the new cabinet or parliament selects 8 individuals to fill those Example: Swaziland.
      • Pros: Provides a bonus to the winning coalition. Ensures that the reserved seat members have some connection and influence with the governing coalition.
      • Cons: Not directly Representatives may be beholden to the ruling party/coalition.
  1. Separate reserved seat lists, general election winner-take-all: Each party produces a list of the eight candidates they would choose to fill the eight The party with the largest vote share in the general election gets to fill the seats using its list.
      • Pros: Voters can take the reserved seat list into account when they cast their vote. Provides a bonus to the winning party. Increased probability that the reserved seat members have some connection and influence with the governing coalition.
      • Cons: Not directly Representatives may be beholden to the largest party. The votes of non-group voters help determine who wins.
  1. Separate reserved seat lists, voters vote directly for their preferred list, winner-take-all. Each party produces a list of the eight candidates they would choose to fill the eight The list of each party appears on the ballot and voters vote for their preferred party reserved seat list. The list with the most votes wins all eight seats.
      • Pros: Voters vote directly for their most preferred list. Encourages party-centered campaigning. Allows voters more choice—they can select one party for the general election, and a different party’s list for the reserved seats if they so choose.
      • Cons: Lengthens the ballot. The votes of non-group voters help determine who wins. Somewhat less likely that the winning list will be connected to the winning party/coalition.
  1. Separate vote for each group of candidates, but all voters can cast a vote: Nominees for each reserved/sectoral category are included on the ballot (so eight separate races), and voters select their most preferred candidate among each of the eight sets of candidates (so voters cast eight separate votes).
      • Pros: Voters vote directly for their most preferred
      • Cons: Significantly lengthens the More information required from voters. Encourages candidate-centered campaigning. The votes of non-group voters help determine who wins.
  1. Separate ballots for each of the eight groups, with only the members of that group being eligible to vote. A separate ballot is produced for each of the eight groups, and voters who register as members of a group would receive the ballot for that group and have a chance to This could be in addition to the regular ballot (as in Pakistan for women’s seats) or voters could be required to choose whether they want to vote in the general election, or in the reserved seat election (as in New Zealand for Maori seats).
      • Pros: Voters vote directly for their most preferred Only group members have a voice in selecting their representative.
      • Cons: More costly, and much more complicated to

This list is certainly not exhaustive—I imagine the drafters can come up with other creative arrangements—but the list is illustrative of the issues and trade-offs that they will have to grapple with as they figure out how to fill these reserved seats.



 Perhaps the most notable innovation in the BOL is the adoption of a parliamentary system. The two key features that distinguish parliamentary from presidential systems are the linked origin and survival of the chief executive and legislature. Under a presidential system the origins and survival of the executive and legislative branches are entirely separate—they are elected separately through popular (usually direct) elections, and they each sit for fixed terms, with their survival in office independent of actions by the other branch (outside of special cases like impeachment). By contrast, in parliamentary systems the chief executive (prime minister/chief minister), and by extension, the government, are selected by the legislature, and to continue in office the chief executive and cabinet must maintain the support (or confidence) of the legislature. If the chief minister loses that support, he or she can be replaced, either through the selection of a new government, or by dissolving parliament and calling a new election. (In many parliamentary systems the chief minister also has the power to unilaterally dissolve parliament and call for new elections—more on this below.) If one party captures a majority of the seats in parliament, government formation is fairly straightforward—that party can form a government on its own. However, if no single party captures a majority, a coalition of parties must come together to agree on a chief minister and government, with cabinet posts shared among members of the governing coalition.

In the case of BARMM, the BOL specifies that the Chief Minister is elected by a majority vote of all members of the Parliament. Parliament has the power to pass a vote of no-confidence against the government (Chief Minister and cabinet) with a two-thirds majority. Upon a successful no- confidence vote the Chief Minister is required to advise the Wali to dissolve parliament and call new parliamentary elections within 120 days. However, there are number of unresolved questions that will need to be addressed in the Electoral Code or other legislation. These include the following:

    • Does the Wali formally appoint a formateur? The formateur is the individual who is given the first chance to form a In many parliamentary systems, this is a formal step in the process. In other cases, there is no formal appointment of a formateur and negotiations to form a new government occur more informally.
    • If there is a formateur, how are they selected? Does the Wali have to appoint the leader of the largest party, or can they select the leader of any party? By convention, the leader of the largest party generally gets the first chance to try and form a government, but this is not always the
    • Is a formal vote of investiture required from parliament (a vote to seat the new government)?
    • Does the Chief Minister have the authority to dissolve parliament and call early elections without first losing a confidence vote? In many parliamentary systems, it is possible for a designated authority (often the prime minister) unilaterally to dissolve parliament and call early elections even in the absence of a confidence vote. Given that the BOL does not explicitly give the Chief Minister this power it would be appear that this is not the case, and this has important implications for the powers of the Chief Minister/government relative to the parliament (more on this below).
    • Are there any limits to votes of confidence? Some parliamentary systems place restrictions on the use of this tool—for example, prohibiting a vote of no confidence early in a government’s term, or restricting the number of times a vote of no-confidence can be called each
    • Does Parliament have to approve cabinet members outside of the Chief Minister and the two Deputies?
    • Does Parliament have the power to remove members of the cabinet without resorting to a vote of confidence?

In addition to working out the details about government formation and dissolution under parliamentarism, drafters will want to give attention to supporting institutions. To return to a theme raised earlier in this document, the nature of the party system is particularly important—both the number of parties, and the strength of those parties. The efficiency, efficacy, and stability of BARMM governments will hinge, in part, on what kind of party system develops. Let’s start with the number of parties.


Number of parties

Given the dependence of the executive on maintaining the support of the legislature, parliamentarism tends to work best when there are a modest number of parties. As the number of parties increases (and the size of the largest parties diminishes) we gain better representation—more parties means that it is more likely that voters can vote for a party that is close to them, one that best represents their interests. But the proliferation of parties also comes with at least three significant costs. First, the process of government formation becomes more difficult as the number of parties rises (witness the repeated government formation drama in Israel). Second, government/cabinet stability tends to decrease as the number of parties in the governing coalition increases (think Italy, with an average government duration of thirteen months since WWII). Third, a large number of parties means that there are more actors who can slow or block policy change—thus the association between party system fragmentation and policy stalemates or gridlock.

The Philippines need not look far to see the risks associated with excessive party system fragmentation under parliamentarism. Party system fragmentation was the norm in nearby Thailand during the 1980s and 1990s. During those two decades the largest parties regularly fell well short of a parliamentary majority—necessitating large, multi-party coalition governments. These governments were notoriously short lived—the average cabinet lasted less than 18 months. By way of comparison, presidential Philippines held four elections for president during the 1980s and 1990s, including two under authoritarian auspices in 1981 and 1986, resulting in four different heads of government (Marcos, Aquino, Ramos, Estrada). Over the same period, Thailand held seven parliamentary elections and was headed by ten different governments (Kriangsak, Prem, Chatichai, Anand, Suchinda, Anand II, Chuan, Banharn, Chavalit, Chuan II).[iii] With such large coalitions and short-lived governments policymaking in Thailand suffered greatly—needed policy reforms were neglected, contributing to Thailand’s vulnerability to the Asian Financial Crisis. When Thailand tackled constitutional reform in 1997, a major focus of reformers was reducing the number of parties.

What do we expect the number of parties to be in BARMM? As discussed above, a lot depends on the choices made by the BTA in the Electoral Code. Most importantly there is the critical decision as to whether political parties can compete in both district and party tiers. (To recap, demarcation of the two tiers can be expected to lead to a ballooning in the number of parties.) A second important decision, also discussed above, is how the district and party tiers may or may not be linked in the allocation of seats (with the more complicated compensatory system likely to increase the total number of parties). Third, there is the choice of whether to impose a threshold, and, if so, at what percentage level.

In addition to these decisions, three other factors could combine to push the party system towards greater fragmentation. First, one particular feature of the system for electing party seats, as already mandated by the BOL, is highly permissive of small parties. This relates to what political scientists call district magnitude, or the number of seats in a district. The more seats in a given electoral district, the greater the number of parties that can compete. Under the BOL, the entire BARMM is constituted as one single electoral district, meaning that 40 seats will be coming from one district. We can expect this to make even very small parties viable contenders for seats in the parliament (Hicken 2019a, 2019b). Second, we know that presidential elections tend to have a deflationary effect on the number of parties in a political system. All else equal, we would thus expect the adoption of parliamentarism to be associated with more parties. Third, the deflationary effect of presidentialism was heightened in the Philippines due to the tradition of party switching— specifically officeholders up and down the political system switching parties to align with the president’s party after the election. Such switching is banned under the BOL, which eliminates the post-election band-wagoning that tended to reduce party system fragmentation in the Philippines. In summary, the removal of the presidentialist “brake” on the number of parties would lead us to expect more parties on balance.[iv]

Other institutional design decisions could also have bearing on the number of parties. First, party regulations can shape the fragmentation of the party system. The more stringent the regulations, the fewer parties will form and compete. Based on the BOL it seems clear that a major goal is to spur the development of regional parties. As noted above, what counts as a regional party is not laid out in the BOL and will need to be defined in subsequent legislation. Drafters need to consider what the requirement for regional parties is meant to imply. One interpretation could be that, in order to be eligible, parties must stand for election only within BARMM—meaning that national parties are prohibited from fielding candidates. Note, such an interpretation would also disqualify BARMM parties that wished to field candidates elsewhere in the Philippines outside of BARMM, including other parts of Mindanao. Another possible way to read the requirement is that parties standing for election must have a verifiable presence in the region. For example, parties could be required to maintain permanent offices across a certain number of the five provinces of BARMM, or to have a certain number of members from each of the BARMM provinces. This would open the door to parties from outside the region being able to compete in elections, provided they were willing to invest in party building within the BARMM. Regardless of how regional parties are defined, requiring parties to maintain a large number of local offices or have a large number of members will tend to reduce the number of parties competing for election.

Second, the BOL is silent on whether there will be geographic distribution requirements. Some countries put in place regulations designed to discourage small, local parties. For example, they may require that parties field candidates in most or all contests across the territory. Or they may require that votes for winning candidates or parties be distributed across the territory. An example of the latter would be Indonesia, which requires presidential candidates to receive at least twenty percent of the vote in at least half of the provinces in order to win in the first round. The presence of distribution requirements would likely reduce the number of parties standing for election.

Finally, rules regarding party finance can shape whether parties form or remain active. The more public funding is available to subsidize parties, the easier it will be for parties to form and compete. I’ll talk more about party funding in the next section.


Party Strength

In addition to the number of parties, the strength of political parties—what we sometimes refer to as their level of institutionalization—is a crucial determinant of how well parliamentary systems function. Parliamentary institutions work best where political parties are strong, cohesive organizations with meaningful policy platforms, and stable roots in society. Traditionally, such parties have been absent in the Philippines, and this is associated with a host of problems.[v] First, needed national policies have been chronically undersupplied. Philippine policy makers are generally unable to transcend the powerful economic interests that have long dominated Philippine politics (Hutchcroft 1998). As a result, public policy caters to the interests of narrow elites at the expense of broader national interests (de Dios and Hutchcroft 2003). Second, weak parties have also meant that Philippine politicians operate with very short time horizons (Lim and Pascual 2001, Hutchcroft and Rocamora 2003), yielding chronic underinvestment or inefficient investment in public services, human capital, and physical infrastructure. Third, weak parties, often the vehicles of powerful personalities, are unable to constrain party leaders, particularly presidents. Hence, Philippine policy has been dependent on the peculiar preferences and personalities of individuals, which, in turn, undermines the predictability and credibility of policy (Balisacan and Hill 2003; de Dios and Esfahani 2001; Hutchcroft 2000). Fourth, the Philippines exhibits high levels of patronage and clientelism (Hicken, Aspinall and Weiss 2019). Finally, the failure of parties to adequately respond to broader societal interests has meant that pressures for reform often take “extra-parliamentary—and even extra-legal—forms…” (Hutchinson 2001: 57), resulting in periodic eruptions of political instability and a concomitant erosion of investor confidence.

This is the foundation on which a new BARMM will be built, and absent institutional innovations, there is little reason to believe that the BARMM party system will look very different. However, institutional reforms can help create an environment more conducive to strong parties. The adoption of parliamentarism is one step in this direction. On average, parties tend to be stronger in parliamentary systems compared to presidential systems. However, while parliamentary systems function best with strong parties, such parties don’t automatically arise under parliamentarism. Other features of the institutional environment, particularly the electoral system, will shape the types of parties that emerge. One need only look to Thailand of the 1980s and 1990s to see that weak parties can exist within parliamentary democracies, undermining government stability and performance. Thai parties during this period were little more than ephemeral electoral alliances of convenience, and with party labels meaningless politicians were largely focused on providing resources to narrow constituencies, with little attention or interest in broader public policy.

So, what kinds of institutions encourage the development of stronger parties? Generally, these are rules that empower party leaders and incentivize voters and politicians to invest in the party label. The BARMM already includes some such provisions, including the choice of a parliamentary system. But if the goal is to encourage stronger parties, much more could be done. One tool that might be considered is the Chief Minister’s power to dissolve parliament and call new elections. As discussed earlier, in the BOL it is unclear whether the Chief Minister can call new elections absent a vote of no- confidence. There would be trade-offs to granting the Chief Minister such power. On the one hand, there is a danger that such power could be abused, leading to political instability. On the other hand, the power to call an early election can be used by government heads to keep party members in-line and hold governing coalitions together. For example, in Thailand, Thaksin Shinawatra was able to use the threat of new elections to prevent factions of his party from jumping ship and joining other Parties.[vi] Prohibiting the Chief Minister from calling early election deprives the government of one of its chief tools for cultivating party discipline and majority cohesion.

Party financing is another tool for party building. If we believe that strong parties are vital to a well- functioning democracy, then there is a case to be made for public investment in political parties, much as the government invests in other crucial utilities. For this reason, dedicating funds to support party building and electioneering have become quite popular in recent years. But as always, there are questions and trade-offs to grapple with regarding the public support of parties. For example, is the support financial (grants to parties) or in kind (free air-time during elections) and what are the pros and cons of each model? Which parties should be eligible for public support? If resources are made available to all registered parties, regardless of size or performance, this can have a number of consequences: a) assuming a fixed pool of resources, the funds each party receives will be less, b) more parties will choose to enter and remain in the political system than might otherwise do so, and, c) parties may be content to remain small rather than trying to build and broaden the base of the party. The alternative is to tie public resources to party size or performance, with larger, more successful parties getting more resources. This has the advantage of investing in the parties that are the most consequential from a political and policy perspective, but this also places at a disadvantage new parties who wish to enter and challenge the status quo.

Thailand’s experience with state funding of political parties underscores some of the challenges involved. Thai scholars have found that public funding is a double-edged sword. While it has supported many new parties, it has also led to a plethora of problems, including parties attempting to boost their funding by setting up fake branches and enrolling fake members, and party leaders using the funds for their personal expenses rather than for building the party (Punchada 2021).

Finally, requiring all candidates and elected officials to belong to and remain in a political party can help generate incentives for politicians to invest in building stronger parties. The BOL is unclear whether candidates have to be a member of a political party in order to stand for election, though references to “unaffiliated” MPs implies that independent candidacies may be allowed. Allowing independents would work against a goal of building stronger parties, but regardless, whether or not party membership is required should be clarified in the Electoral Code.

The BOL does place new restrictions on party switching, a first for the Philippines. Under the BOL MPs who switch parties must give up their seat (which should eliminate post-election party hopping), and candidates must be a member of a party for at least 6 months prior to an election (which should reduce pre-election turncoatism). However, the law is unclear about the following questions:

    • If an MP leaves the party, but does not join another party (i.e. becomes an independent) do they lose their seat?
    • If an independent/unaffiliated MP joins a party, does that constitute party switching?
    • If an MP is forcefully expelled from a party, do they lose their seat?
    • If a party splits into two different parties, or two parties decide to merge and form one party, does that count as party switching for the members of those parties?
    • If two parties decide to merge, but an affiliated MP doesn’t wish to be part of the new party and opts to leave, do they lose their seat?



I am sometimes asked what the “best” institution is for a given country. My answer, while perhaps not very satisfying, is always the same: “It depends.” It depends on what the goals and priorities are of a given country, as well as the broader political and institutional context. As the drafters of the Electoral Code and other laws continue their work, and as societal actors work to give their input to the drafting process, I invite us to return to the three questions raised at the beginning of this paper. What are the goals we are trying to achieve or what problems are we trying to solve? What do we think are the causes of the problems we would like to solve? How would our proposed institutions help accomplish those goals or address the problems we’ve identified? Answering those questions is a crucial first step. We should also keep in mind the three principles described earlier: a) institutions embody trade-offs, b) institutional reforms do not take place in a vacuum so be prepared for unexpected consequences, and c) keep it simple. One arena where these trade-offs is most manifest is in the party system. The party system will serve as the foundation of BARMM’s parliamentary structure of government. Given the crucial role it will play in shaping executive authority and, ultimately, the extent to which BARMM governments are stable and effective, drafters would do well to consider carefully how institutional design choices will affect the development and nature of the BARMM party system.

Allen Hicken is professor of political science and a research professor at the Center for Political Studies and the Center for Southeast Asian Studies at the University of Michigan. The views and opinions expressed in this article are those of the author and are not meant to represent those of the Australian National University, the Australian Government, Access Bangsamoro, its proponents, or affiliates.

Appendix: Tradeoffs in the Design of Electoral System[vii]

Institutions embody trade-offs—choices about which goals we want to prioritize and which we are willing to sacrifice. For example, one of the most common institutional tradeoffs is that between decisiveness and resoluteness (MacIntyre 2002). Both goals are laudable. We want governments that can make decisions in a timely, efficient manner, responding quickly to challenges and opportunities (decisiveness). At the same time, we value policy stability, and the predictability and credibility it brings (resoluteness). However, these two goals are in tension with each other. Institutions that centralize political power trade-off resoluteness for greater decisiveness, while those institutions that disperse power sacrifice some decisiveness for greater policy stability.

We see similar trade-offs in the choice of electoral system. Specifically, we can think of these trade- offs as falling along two dimensions—an interparty dimension and an intraparty dimension (see Figure 1) (Shugart 2001). Along the interparty dimension is the trade-off between what we might call plurality/majoritarian visions of democracy versus proportional/representation visions of democracy.[viii]

    • The plurality/majoritarian vision places great value on decisiveness and effectiveness— governments being able to pass policies quickly and It also values identifiability (voters being able to vote directly on who they want to control government) and accountability and clarity of responsibility (being able to assign clear credit or blame to policy makers in order to hold them accountable). This corresponds with a preference for a small number of large parties—ideally two parties or party blocks—one of which will likely gain full control of the government.
    • The proportional/representation vision (also called the power-sharing or consensus vision), by contrast, values giving voice and representation to lots of groups and A variety of interests should have seats at the policy table. The proportional vision values power-sharing, inclusiveness, and consensus over majority rule, and policy stability/credibility over decisiveness. This corresponds to a preference for lots of parties and coalition governments.

Again, a tradeoff exists between these two visions. Moving in a plurality/majoritarian direction can come at the cost of representation and policy stability, while moving in a proportional direction sacrifices some decisiveness and clarity of responsibility.

Along the second, intraparty dimension we have a tradeoff between personalistic and party-centered politics.

    • Personalized systems value close ties between individual politicians and local voters/constituents. They want politicians with whom voters can personally interact and to whom they can go for The focus, under this vision, is on incentivizing politicians to respond to local needs and local interests. This corresponds with a preference for candidate- centered electoral systems and weaker, decentralized parties.
    • The party-centered vision, by contrast, privileges close ties between politicians and their party. Under this vision politicians’ primary loyalty should be to party leaders and the package of policies the party label represents. The focus of party-centered systems is on policies as defined by party leaders and on responding to broad constituencies, with voters being able to clearly distinguish between parties and hold parties collectively accountable. This is reflected in a preference for party-centered electoral systems and strong, centralized parties.

Again, there is a trade-off here. As we move toward more personalized systems, we sacrifice collective partisan accountability and a broader policy focus, while movement in the opposite direction comes at the cost of connections between politicians and local constituencies/voters.

The bottom line is that there is no ideal system, as different actors will place different weights on these competing values. However, we do know that as polities approach the extremes of these dimensions, they are more likely to experience systemic failure–what Shugart (2001) calls electoral system pathologies or inefficiencies (see also Hicken 2019b for more details). Figure 1 presents these extremes across the interparty and intraparty dimensions. At the extremes of the interparty dimension (horizontal axis) we get a large number of parties contributing to government instability, policy deadlock, and low identifiability and accountability (hyper-representativeness), and the at other extreme, electoral systems produce governments representing well under a majority of the electorate (pluralitarianism) (Shugart 2001). Turning our attention to the intraparty dimension (vertical axis), at one extreme narrow political interests completely swamp collective party and national interests (hyper-personalism or hyper-particularism). In hyper-personalized systems we see resources flowing to narrow interests while broader, collective interests are neglected. At the other extreme we get hyper-partisanship, where the connection between individual legislators and voters/local constituents is completely broken. Politicians are focused solely on pleasing the party leadership and are not responsive at all to local constituents and their interests (ibid).

When countries find themselves at any of these extremes this can be the trigger for institutional reform. To illustrate, Figure 1 shows some prominent examples of where countries were located prior to their electoral system reform efforts. New Zealand’s system regularly produced manufactured majorities—governments headed by single parties that had the support of less than half of the electorate. At the other extreme, the Italian and Israeli electoral systems produced multi- party governments that were unstable and ineffective. Japan’s SNTV electoral system was the epitome of a hyper-personalized electoral system, and Venezuela’s hyper-partisan electoral system contributed to an acute representation and accountability crisis in that country. Finally, Thailand’s pre-reform system scored high both in terms of hyper-personalism and in terms of hyper- representativeness. But where does the Philippines sit?

On the intra-party dimension, one can make the case that the Philippines is on the hyper- personalized (lower) side of the continuum. Weak political parties are, in fact, a defining feature of Philippine politics. Politicians are almost entirely focused on cultivating their narrow political bailiwicks at the expense of party loyalty and a broader, national focus. In terms of the inter-party dimension, I would place the Philippines on the pluralitarian (right-hand) side of the continuum. After martial law the Philippines did not return to the pre-1972 system of two major national parties. With the introduction of the party list the party system further fragmented, producing legislatures with a large number of parties. At the same time, however, presidents are effectively assured (after what is commonly a rampant post-election process of party switching) of gaining a legislative majority. The only mystery regarding the government coalitions in the Philippines is which presidential candidate will end up leading them. Once the election is over, party switching to the winning candidate’s party will typically give the president a large majority.

Figure 1: Inter- and Intra-Party Dimensions of Electoral Systems

(Adapted from Shugart 2001 and Hicken 2019b)

Notes: The figure presents the positions of the following countries prior to electoral reform: Israel (ISR), Venezuela (VNZ), New Zealand (NZ), Thailand (TH), Italy (ITL), and Japan (JP)—all of which have fairly recently undertaken electoral reform. In addition, it includes the current position of the Philippines (PHL).




Aspinall, Edward. 2019. “Lessons from a Neighbor: The Negative Consequences of Indonesia’s Shift to the Open List.” In Strong Patronage, Weak Parties: The Case for Electoral Redesign in the Philippines, edited by Paul D. Hutchcroft, 93–108. Mandaluyong City, Philippines: Anvil.

Balisacan, Arsenio N. and Hal Hill. 2003. “An Introduction to the Key Issues.” In Arsenio M. Balisacan, and Hal Hill, eds. The Philippines Economy: Development, Policies, and Challenges. Oxford University Press.

de Dios, Emmanuel, Hadi Salehi Esfahani. 2001. “Centralization, Political Turnover, and Investment in the Philippines.” In J. Edgardo Campos, ed. Corruption: The Boom and Bust of East Asia. Ateneo de Manila University Press.

Hicken, Allen. 2019a. “Electoral System Design: Why it Matters to Development Outcomes.” In Strong Patronage, Weak Parties: The Case for Electoral System Redesign in the Philippines, edited by Paul D. Hutchcroft, 27-40. Mandaluyong City, Philippines: Anvil Publishing.

Hicken, Allen. 2019b. “When Does Electoral System Reform Occur?” In Strong Patronage, Weak Parties: The Case for Electoral System Redesign in the Philippines, edited by Paul D. Hutchcroft, 59-95. Mandaluyong City, Philippines: Anvil Publishing.

Hicken, Allen, Edward Aspinall, and Meredith Weiss, eds. 2019. Electoral Dynamics in the Philippines: Money Politics, Patronage, and Clientelism at the Grassroots. Singapore: NUS Press.

Hutchcroft, Paul D. 1998. Booty Capitalism: The Politics of Banking in the Philippines. Ithaca: Cornell University Press.

Hutchcroft, Paul. 2000. “Obstructive Corruption: The Politics of Privilege in the Philippines.” In Mushtaq H. Khan and Jomo K.S., eds. Rents, Rent-Seeking and Economic Development: Theory and Evidence in Asia. Cambridge: Cambridge University Press.

Hutchcroft, Paul, ed. 2019. Strong Patronage, Weak Parties: The Case for Electoral System Redesign in the Philippines. Mandaluyong City, Philippines: Anvil Publishing.

Hutchcroft, Paul, Joel Rocamora. 2003. “Strong Demands and Weak Institutions: The Origins and Evolution of the Democratic Deficit in the Philippines.” Journal of East Asian Studies 3(2):259- 292.

Hutchinson, Jane. 2001. “Crisis and Change in the Philippines.” In Garry Rodan, Kevin Hewison and Richard Robison, eds. The Political Economy of Southeast Asia. Oxford: Oxford University Press.

Lim, Joseph Y., Clarence G. Pascual. 2001. “The Detrimental Role of Biased Policies: Framework and Case studies.” The Political Economy of Corruption: Studies in Transparency and Accountable Governance No. 3, Transparent and Accountable Governance (TAG) Project, Manila.

MacIntyre, Andrew J. 2002. The Power of Institutions: Political Architecture and Governance. Cornell: Cornell University Press.

Punchada Sirivunnabood. 2021. “Thailand’s Malfunctioning Political Party System.” Perspective. ISEAS Yusof Ishak Institute. Issue 2021:76.

Shugart, Matthew Soberg. 2001. “Electoral “Efficiency” and the Move to Mixed-member Systems.” Electoral Studies 20.2 (2001): 173-93.

Teehankee, Julio C. 2019. “Untangling the Party List System.” In Strong Patronage, Weak Parties: The Case for Electoral Redesign in the Philippines, edited by Paul D. Hutchcroft, 151–67. Mandaluyong City, Philippines: Anvil.




[i] Article 7, Section 8 also raises the possibility that electing representatives could vary depending on the group involved, requiring adherence to “customary laws and processes” when electing non-Moro indigenous representatives.

[ii] For a more detailed discussion of two of the most common trade-offs in institutional design, majoritarian v. proportional and party-centered v. candidate-centered, see the Appendix.

[iii] The Anand governments were appointed.

[iv] That said, the BARMM party system would be even more fragmented if there had been the adoption a party list seat cap similar to the one used at the national level. The national three-seat cap has artificially inflated the number of parties that win seats in the party list system. This makes the system inherently non-proportional in character, as the strongest party list party could hypothetically win 50, 60, or 75 percent of the votes and still be limited to a mere three seats. Fortunately, the BOL aims to create a genuinely proportional system, and does not include any provision for a cap on the number of seats; a party winning 50 percent of the votes should receive roughly 50 percent of the PR seats (i.e., 50 percent of the 40 seats allocated to PR, or 20 seats total).

[v] This paragraph draws on Hicken 2019a.

[vi] Similar to the BOL provisions, Thai MPs had to be members of parties for a period of time before becoming eligible to run for election. If they switched the prime minister could immediately call new elections, leaving them ineligible for elections. Note that Thaksin was able to use his increased control of party factions to build a secure parliamentary majority that effectively shut out his political opponents.

[vii] This section borrows from Hicken 2019b.

[viii] Majoritarian as used here refers to the general category of electoral systems where the winner “takes all.” Within this larger category are lots of different kinds of arrangements, including plurality rule, which is widely used in the Philippines, and a majority run-off system, as used in some other countries.