INTRODUCTION
As countries in Asia have democratized over the last twenty-five years, claims have arisen that politics in the region has become more judicialized (Dressel Reference Dressel2012)—a trend well documented in other parts of the world (Tate and Vallinder Reference Tate and Vallinder1995; Sieder, Schjolden, and Angell Reference Sieder, Schjolden and Angell2005). The increasing political engagement of courts has thus raised questions about their independence. There are perceptions that appointments to the high court may create a sense of obligation, with consequences for judicial decision making, particularly in high-profile political cases. The “loyalty effects” discussed in the US context by Epstein and Posner (Reference Epstein and Posner2016) are a growing concern for Asia’s young democracies, where populist leadership and executive assertiveness are increasingly testing constitutional checks and balances (Kenny Reference Kenny2018; Bünte and Thompson Reference Bünte and Thompson2018; Croissant and Haynes Reference Croissant and Haynes2021).
The Supreme Court of the Philippines (SC) illustrates many of these concerns. A central institution in Asia’s oldest democracy, the SC was equipped with far-reaching review powers after democratic institutions were restored in 1986. It has since been called upon regularly to resolve high-profile political cases in such diverse areas as economic policy, human rights, separation of powers, and executive prerogatives (Pangalangan Reference Pangalangan and Albert2014). Combined with its regular engagement in contested electoral matters, the SC is considered unusually activist for the region (Tate Reference Tate1994; Dressel Reference Dressel, Bünte and Dressel2017).
Although such high-level engagement makes it more politically vulnerable, as was the constitutional court in Thailand, the Philippine SC remains a respected institution, traditionally enjoying positive trust ratings in public polls.Footnote 1 Nevertheless, scandals and controversial decisions have also caused fluctuations over time in public approval (Deinla Reference Deinla2014). There is a perception that its justices are susceptible to partisanship due to a sense of loyalty born out of gratitude for their appointment to the Court (Agabin Reference Agabin2012, 21–22). Meanwhile, the 2012 impeachment of Chief Justice (CJ) Renato Corona and the 2018 removal of CJ Sereno via a quo warranto writ are reminders that the SC is vulnerable to executive pressures (Dressel and Inoue Reference Dressel and Inoue2018).
Concerns over the erosion of judicial independence have reached new heights under President Duterte (2016–22). Emboldened by high public approval ratings, Duterte has undermined the post-Marcos liberal constitutional order through both legal and extralegal means (Dressel and Bonoan Reference Dressel and Bonoan2019). This heightened scrutiny of the SC’s willingness to defend the constitutional order after the declaration of martial law in Mindanao (2017–20), widespread human rights violations during the “war on drugs,”Footnote 2 and a spike in killings of lawyers and judges.Footnote 3 Combined with an unprecedented number of presidential appointees to the SC and a string of high-profile decisions that appear to favor the Duterte administration, it is not surprising that public concerns about the SC’s independence have been mounting (Gatmaytan Reference Gatmaytan2018; Ibarra Reference Ibarra2020; Fonbuena, Gajet Buenafe, and Teodoro Reference Fonbuena, Gajet Buenafe and Teodoro2021).
In light of these concerns, we asked the following questions: Other than the law, what factors might influence the decisions of judges in high-profile cases? Do loyalty effects play a role in these cases? Also, does the length of time serving as a justice affect these loyalty dynamics? These questions are particularly relevant given the erosion of other democratic institutions. Moreover, by the time Duterte’s term ended in 2022, he had appointed all but two of the members of the SC.
Compared to its peers in the region, the Philippine SC has been relatively well studied but conclusions about what influences judicial behavior have been far from clear. Empirical studies have drawn attention to the sociobiographic backgrounds of judges (Tate Reference Tate1970; Gatmaytan and Magno Reference Gatmaytan and Magno2011); tested for the impact of regime variables on court performance (Tate and Haynie Reference Tate and Haynie1993, Reference Tate and Haynie1994); and provided measures of ideological points for justices (Pellegrina, Escresa, and Garoupa Reference Pellegrina, Escresa and Garoupa2014). Attitudinal and principal-agent-based models have suggested support for alignment of the performance between individual justices and the interests of the presidents who appointed them (Escresa and Garoupa Reference Escresa and Garoupa2012, Reference Escresa and Garoupa2013; Desierto Reference Desierto, Schofield and Caballero2015).
On the other hand, qualitative studies have offered rich narratives of the transformation of the SC and the dynamics underpinning how it functions from political-institutional (Agabin Reference Agabin2012), historical-legal (Cruz and Cruz-Datu Reference Cruz and Cruz-Datu2000), and investigative (Vitug Reference Vitug2010, Reference Vitug2012) perspectives. Most of these accounts explicitly acknowledge the importance of informal dynamics, such as presidential influence on nominations and court stacking; the personal characteristics that shape judicial leadership; or patterns of obligation, friendship, and loyalty to political actors that influence how justices decide cases. Nevertheless, evidence to support these dynamics has been largely anecdotal (although see Dressel and Inoue Reference Dressel and Inoue2018).
Here we draw on broader comparative judicial politics studies of loyalty effects in the judiciary (Epstein and Posner Reference Epstein and Posner2016; Leslie, Robinson, and Smyth Reference Leslie, Robinson and Smyth2021) and the pioneering empirical work of Escresa and Garoupa on the SC (Escresa and Garoupa Reference Escresa and Garoupa2012, Reference Escresa and Garoupa2013). As highlighted by numerous studies in the fields of economics, political science, and psychology, the dynamics of gratitude, obligation, and reciprocity that are born out of personal or professional ties can all result in relational ties that can override professional or ideological considerations (Finan and Schechter Reference Finan and Schechter2012; Baldwin Reference Baldwin2013; Cruz, Labonne, and Querubín Reference Cruz, Labonne and Querubín2017; Leslie, Robinson, and Smyth Reference Leslie, Robinson and Smyth2021). It has been suggested that in the United States, justices may be willing to vote for the appointer out of a “sense of personal obligation” or gratitude (Scigliano Reference Scigliano1971, 132; Epstein and Posner Reference Epstein and Posner2016, 408). Not surprisingly, such dynamics are even more intense within judicial systems dominated by patronage rather than merit-based appointments (Sanchez Urribarri Reference Urribarri2010; Hausegger, Riddell, and Hennigar Reference Hausegger, Riddell and Hennigar2013; Epstein and Posner Reference Epstein and Posner2016).
Extending such debates to the Philippines has much to offer not only for studies of the Philippine SC but also for broader comparative judicial politics scholarship. We hypothesize that executive influence and loyalty affect judicial decision making in highly political cases, given that ideological preferences can be hard to discern in the fluid political context of the Philippines. Moreover, the unique single presidential term in the Philippines also offers new dynamics of interest to the comparative judicial behavior literature. Our study suggests that loyalty effects are mitigated by strategic considerations, as expressed at the end of the presidential term, but that it may be possible for presidential loyalties to extend beyond the term of the appointer when incoming presidents form alliances with their predecessors, as is suggested in factional voting behavior observed on the bench.
To test for these assumptions, we use an original data set based on a three-step methodology for identifying megapolitical cases from 1987 through 2020. Following Ran Hirschl, we define megapolitical cases as those that go beyond issues of procedural justice and political salience to include “core political controversies that define (and often divide) whole polities” (Hirschl Reference Hirschl2006, 725). As we outline in detail below, we identified seventy cases from 1987 to 2020 and constructed sociobiographic profiles of the eighty-six SC justices who served during this period. We then tested patterns of judicial alignment and dissent in the seventy cases to explore the extent to which members of the Court have taken a counter-majoritarian role in the political system.
First, our analysis of descriptive data and the results of the regression analyses for our high-profile cases for our sampling period (1987–2020) lend support to beliefs that appointees of a sitting administration are more likely to vote for that administration. Second, justices are more likely to vote for the administration in the first year of their appointment. Finally, a closer look at the cases decided under the Duterte administration also reveals patterns not previously considered in the literature. Among them, votes in favor of the government have increased and loyalties illustrated in the data set have become more pronounced under Duterte. Equally interesting, the voting patterns of cohorts of presidential appointees demonstrate broader factional alignments among SC justices. For instance, not only are Duterte appointees regularly voting together but, particularly early in his presidency, their votes seem to be aligned with the votes of justices appointed by President Gloria Macapagal Arroyo (2001–10). The alignment votes thus regularly left in the minority the appointees of President Aquino III (2010–16). This apparent alignment explains high dissent rates but also suggests a pattern of factional alliances among justices based on presidential administrations. Rather than ideological affinities, these alignments raise the question of how deeply the SC itself might be embedded in the political dynamics of the Philippines.
To illustrate the argument, the article is structured as follows: First we address the SC’s institutional background and historical performance with particular attention to the nomination process. We then briefly summarize the theories of judicial behavior and the guiding hypothesis of the study. After the data set is described, we discuss the empirical results, with particular attention to the Duterte period, and then articulate our conclusions.
SUPREME COURT: ESTABLISHMENT, POWERS, AND PERFORMANCE
The Philippine SC has been central to the country’s political system since it was established in 1901. Modeled after the US judicial system,Footnote 4 it became part of the Commonwealth of the Philippines in 1935. It became the top court when independence in 1946 abolished the appellate functions of the US Supreme Court. Although its role is occasionally contested, the SC has been regularly called upon to review cases with policy consequences. During the authoritarian government of President Ferdinand Marcos (1965–86), it was generally viewed as having succumbed to the executive, which prompted President Corazon Aquino (1986–92) to repopulate the bench upon the return to democracy.
Under the 1987 Constitution, the SC is at the apex of the Philippine court system with oversight of (1) the three levels of courts of criminal and civil jurisdiction; (2) the tax courts; (3) the special antigraft court; and (4) quasi-judicial agencies. The SC and lower courts are vested with the power of judicial review. The SC has both original and appellate jurisdiction and administrative supervision over all courts and court personnel.
As a direct reaction to the Martial Law period (1972–81), the 1987 Constitution expanded SC review powers (article VIII, section 1). Safeguards for judicial independence have also been reinforced (article VIII, sections 2–6). Judges enjoy security of tenure (article VIII, sections 10–11). The salaries of judges may not be reduced; the judiciary was granted fiscal autonomy and the legislature may not reduce its budget below the appropriation for the previous year (article VIII, section 3). Nor may Congress deprive the SC of jurisdiction over cases enumerated in section 5 of article VIII. Furthermore, the SC supervises all lower courts, with the power to discipline all judges and lawyers (article VIII, section 6). It appoints all court officials and employees in accordance with the Civil Service Law (article VIII, section 5(6)).
The SC is composed of a Chief Justice (CJ) and fourteen Associate Justices, whom the president appoints from a shortlist of candidates provided by the independent Judicial and Bar Council (JBC). The JBC, established by the 1987 Constitution, was meant to establish a merit-based appointment system that insulates judicial appointments from traditional politics. Yet executive influence remains given that the president appoints some members of the JBC. This includes the secretary of justice, who is an ex officio member, and the four regular members of the JBC (a law professor, a retired SC justice, a member of the Integrated Bar of the Philippines, and a representative of the private sector).
Qualifications for the bench are stringent: SC justices must be at least forty years old and have at least fifteen years of experience as a lower court judge or a practicing lawyer (article VIII, section 7). Most SC justices have been judges, although academic appointments are common (see the fifth section, below). Unless they resign earlier or are impeached, justices must retire at age seventy.
The SC may sit en banc or in groups of three, five, or seven justices (article VIII, section 4(1)). En banc cases are decided by a majority of the justices who took part in and voted based on the deliberations (article VIII, section 4(2)). Separate opinions are allowed, but the majority opinion that resolves the case is written by an assigned judge, the ponente.
Compared to other high courts in Asia, the Philippine SC accepts a larger number and a wider range of cases. Although the caseload dropped from 11,810 in 2006 to 3,778 in 2020 (see Figure 1), de-clogging its docket remains a challenge, partly because the number of cases disposed of annually has dropped from 5,302 cases in 2006 to 3,605 in 2020. Thus, with 8,818 cases pending in 2020, the SC carries forward a backlog each year. Backlogs and undue delays exemplify the challenges confronting every court in the country—ranging from low functional efficiency to limited access to justice for the poor, despite decades of donor-supported judicial reform projects.Footnote 5
Supreme Court Caseload, 2006–20.
Source: Judiciary Annual Reports.
Although SC decisions are still of high quality despite the demanding workload, scandals have jeopardized the Court’s reputation and stability. Justices have been threatened with impeachment, although only CJ Corona was successfully impeached. More recently, the majority of SC members voted via a quo warranto writ to remove CJ Sereno (2012–18). Some observers saw the move as not only legally problematic but possibly part of a Duterte agenda to remove one of his staunchest critics on the bench.Footnote 6
Public concerns over executive influence have plagued the SC, particularly in highly political decisions (Gatmaytan Reference Gatmaytan2015). These concerns have intensified under the Duterte administration (2016–22), given the high number of controversial cases decided in favor of the governmentFootnote 7 as well as removal, early resignations, retirements, and short-term appointments of justices since he was elected (see Table 3, below).
THEORY AND HYPOTHESIS
The literature on judicial behavior has identified variables that enter into decision making by high courts (see overview by Baum Reference Baum2006; Roux Reference Roux2018). Personal attributes and attitudes matter (e.g., policy and outcome preferences). Interactions within the bench may also play a role (natural pressure for consensus; concern for court reputation; a common desire to empower the court over competing political and judicial actors). Party politics may have some influence, including loyalty to the appointer. Finally, these variables interact within specific constitutional and doctrinal contexts, some with more, some with less legal formalism.
The theory a scholar follows affects the relative importance of these variables. For instance, the legal model assumes that judges decide in conformity with laws and precedent (Bailey and Maltzman Reference Bailey and Maltzman2011). It supports an image of judges as neutral and apolitical, using technical interpretation to ascertain the law that best applies to a given case (Shapiro Reference Shapiro1981). Attitudinal theorists argue, however, that ideological positions and policy preferences shape judicial decisions (Segal and Spaeth Reference Segal and Spaeth1993, Reference Segal and Spaeth2002). They downplay the influence of the letter of the law and see judges as focused on legal policy. The strategic model of judicial decision making, also guided by the notion of judicial policy preferences, acknowledges that judges take into account the views of other actors and the institutional context—and may even deviate from a preferred outcome to do so (Epstein and Knight Reference Epstein and Knight2013; Spiller and Gely Reference Spiller, Gely, Keith, Daniel and Gregory2010).
A full discussion of these theories is beyond the scope of this article (see, for a comprehensive overview of the literature, Baum Reference Baum1994; Whittington Reference Whittington2000). What is clear, though, is that while different models have increasingly incorporated ideas from each other and widened their scope, capturing dimensions such as loyalty effects has been difficult because the loyalty is to a person (e.g., the appointer) rather than an organized group (e.g., parties) or an idea (e.g., ideology); nor is the propensity to make sacrifices to the benefit of another person a purely rational, strategic undertaking.
In fact, there is growing acknowledgment of potential loyalty effects—whether anchored in a quid pro quo between actors, a cultural sense of reciprocity, or a psychological notion of gratitude that compels the beneficiary to reciprocate (Epstein and Posner Reference Epstein and Posner2016, 408). These dynamics add a layer of constraints on judicial behavior that are not considered in traditional attitudinal or even strategic models used to explain the success of the executive in high-profile political cases.
If loyalty effects are at play in the US Supreme Court, what about in settings where institutions are weaker and judges remain embedded in personalized politics and strong informal networks?
Recent academic debates have raised concerns about how well certain models travel beyond the West (Roux Reference Roux2015; Dressel, Sanchez Urribarri, and Stroh Reference Dressel, Sanchez Urribarri and Stroh2018). In many transitional economies, institutions are weaker and clientelistic, highly relational, and ideological faultlines are difficult to determine—or irrelevant. Thus whether for emotional-psychological or sociocultural reasons, judges might be influenced by loyalty to a particular political patron or a network of relational alliances. While this does not preclude judges from strategic considerations, it raises the question of whether and to what extent loyalties to the appointer are maintained, changed, or refuted as judges navigate high-profile cases.
In response, we propose a model that has been loosely identified as strategic.Footnote 8 Taking inspiration from the work of Epstein and Posner on the US Supreme Court (2016) and Escresa and Garoupa (Reference Escresa and Garoupa2013, Reference Escresa and Garoupa2012) on the Philippine Supreme Court, we use the model to test some critiques of the behavior of judges—how executive appointments might have shaped voting in megapolitical cases, possible potential “loyalty effects,” and whether the length of time they have served on the bench has any effect.
We test for three hypotheses that relate to a strategic understanding of how SC justices behave:
H1. Judges deciding cases during the administration of the president who appointed them are more likely to vote for the government in high-profile cases than those appointed by previous presidents. This seeks to capture the loyalty and strategic alignment to the appointer that many scholars have acknowledged (Epstein and Posner Reference Epstein and Posner2016), which in the context of the Philippines might be reinforced by a cultural sense of obligation and gratitude (utang na loob) and of group conformity (pakikisama) often described in the literature (Agabin Reference Agabin2012, 22).
H2. Judges are more likely to vote for the administration in the first year on the bench than later in their tenure. This captures the “freshman effect” described in studies of the US Supreme Court (Bowen and Scheb Reference Bowen and Scheb1993; Hagle Reference Hagle1993). Reasons cited for this include bewilderment, disorientation, and peer pressure, but for the Philippines, we also posit more deference earlier in the term based on a cultural sense of gratitude (see H1).
H3. There are distinct differences in behavior among presidential appointees. This tests claims by court observers and academics that the justices appointed by Duterte are more likely to vote for the administration than previous generations of appointees (Ibarra Reference Ibarra2020; Gatmaytan Reference Gatmaytan2020).
Our first step is to provide statistics to describe the SC bench over time. Then we look more closely at the voting behavior of individual justices, supported by inferential statistics on how personal traits may account for individual votes in the cases sampled.
DATA SET
We analyzed decisions promulgated by the Philippine SC from 1986 to 2020 (see Table A3) to identify cases that were megapolitical. Drawing on a case selection methodology developed by Kapiszewski (Reference Kapiszewski2011), we selected our cases in three steps based on (1) coverage on the first two pages in two major news articles (e.g., in Philippines Inquirer, Philippines Star);Footnote 9 (2) citations in articles about the SC in law publications (e.g., constitutional law commentary, academic books, law journals, and constitutional law syllabi);Footnote 10 and (3) vetting by leading local constitutional law experts both in terms of initial identification and confirmation of final case list.Footnote 11 This three-stage triangulation process allowed us to (a) comprehensively capture the universe of cases, and (b) narrow them down to those megapolitical cases most aligned with the framework suggested by Hirschl (2008).
While only a small fraction of cases decided by the Supreme Court, these cases are of special interest to us as we can reasonably expect that personal and political factors would matter more to decisions in megapolitical cases in part because of the political high stakes as well as the lack of long-standing legal doctrine for decisions on novel matters.
The individual votes of each justice in the seventy cases constitute 940 observations. The outcome of interest, the dependent variable in the regression, is a vote against the incumbent administration. From the sociobiographic data on the eighty-six justices appointed between 1986 to 2020 who voted in these cases, we identify details such as time on the bench, university affiliation, gender, professional career, and previous workplace.
FINDINGS
We begin by using descriptive statistics to demonstrate the evolution of the bench over the sample period. We then analyze individual voting patterns by case types, Chief Justice tenure, and presidential term. Finally, we apply regression analyses.
The Bench
The sample period coincides with the administrations of Presidents Corazon Aquino (1986–92); Fidel V. Ramos (1992–98); Joseph Estrada (1998–2001); Gloria Macapagal Arroyo (2001–10); Benigno Aquino III (2010–16); and Rodrigo Duterte (2016–22). Of the eighty-six justices the majority were appointed after 1986 during Corazon Aquino’s renewal of the bench and Macapagal Arroyo’s extended term (Table 1).Footnote 12
The Philippine Supreme Court Bench, 1987–2020
Source: Data from SC records. Due to rounding, the totals may not add up to 100%. For universities, the Other category includes: Ateneo de Davao, Francisco Law School, Lyceum, Quezon, San Carlos, University of the East, and University of the Visayas. For prior position, the Other category includes the Commission on Elections (COMELEC) and the legislature (House of Representatives and Senate).
SC membership has not been diverse. For instance, while women were appointed to the bench early (unlike in peer courts in Asia) on the SC, there have been far fewer women than men. Similarly, although the Philippines has an estimated 105 law schools, alumni of only twelve schools have been appointed to the SC. The University of the Philippines and the Ateneo de Manila University account for about 61 percent (fifty-two of eighty-six) of the justices who have served on the Supreme Court since 1986. Relative to population, justices from Luzon are overrepresented (79 percent) and those from Mindanao are severely underrepresented (6 percent)—an illustration of the traditional bias of elite socialization in the capital, Manila, and surrounding Luzon. Most justices have been appointed from within the judiciary (67 percent), followed by the private sector (14 percent), academe (9 percent), and the executive branch (7 percent).
The backgrounds of appointees by administrations can vary considerably, which hints at how much each president’s preferences might influence appointments. There is a recent rise to the bench of San Beda law graduates from an average of 10 percent pre-Duterte to 40 percent during his presidency. Duterte graduated from San Beda. There is a similar increase of appointees from Mindanao and the Visayas—regions closely tied to his power base—from an average of 18 percent pre-Duterte to 33 percent currently. Moreover, all his appointees have been career judges, who dominate the SC bench, with the majority in the past being elevated from the Court of Appeals (79 percent) and the specialized Sandiganbayan Anti-Corruption Court (10 percent).
The composition of the bench raises questions about how successful the institutional arrangements are for both the nomination procedures of the Judicial and Bar Council (JBC) and the final presidential appointment. The JBC was created by the 1987 Constitution to insulate judicial appointments from politics and ensure that appointees are of proven competence, probity, and independence; previously, justices were confirmed by the congressional commission on appointments, which was seen as favoring judges who had a backer (padrino) in the commission. The JBC was also the result of concerns about the unrestrained presidential discretion of Marcos after Congress was abolished in 1973 (Vitug Reference Vitug2010, 108).
However, considering how traditional informal channels may still influence judicial appointments, it is questionable whether the JBC has operated as ideally intended. For instance, not only does the executive still appoint most of the JBC members, but justices appear to be recruited from a shallow pool of candidates with often overlapping social circles and professional trajectories (Gatmaytan and Magno Reference Gatmaytan and Magno2011).Footnote 13 The result is politicized nomination and appointment processes that often lack transparency despite recent efforts by the JBC for greater transparency and public engagement.
Nevertheless, the bench has been relatively stable, with an average tenure of 7.23 years over the sample period. The overwhelming majority of justices retire at seventy, although there are rare resignations (e.g., for medical reasons) and dismissals (Table 2).
Reasons for Leaving the Bench, 1986–2020
Note: Up to three days earlier than the mandatory date of retirement is classified as a normal retirement. Resignation includes early retirement for unknown reasons (7), reappointments, and running for other offices (3).
There are substantial differences in judicial tenure by presidential administrations, and a slight tendency to appoint judges for longer terms toward the end of the presidential term. The trend is most evident under Duterte. Reinforced by a number of resignations, the estimated average term of a justice appointed by Duterte in the past two years was 6.39 years or almost double the 3.82-year average tenure of appointees in the first two years of his administration.
The appointments and turnovers might suggest that the president has some strategic considerations. Since the removal of Sereno in 2018, Duterte has appointed four CJs; his first appointee served less than two months. This revolving door policy extends to appointments of Associate Justices, with Duterte appointing relatively older justices with short tenures especially in the first two years of his presidency (Fonbuena, Gajet Buenafe, and Teodoro Reference Fonbuena, Gajet Buenafe and Teodoro2021). There have also been an unprecedented number of early retirements under Duterte, with one CJ and three Associate Justices retiring before the mandatory age of seventy, much to the bewilderment of observers.Footnote 14 The fast turnover of SC membership since 2016 has resulted in a lack of continuity in leadership and likely contributed to inefficiencies and delays.
Voting Patterns, 1986–2020
The number of megapolitical cases has risen gradually over time to peak in 2019; these often involve disputes between branches of government, as reflected in the distribution of cases by category. Almost a third involved issues of separation of powers (31 percent), followed by executive prerogatives (23 percent) and the bill of rights (19 percent). The remaining 27 percent dealt with economic matters (11 percent), elections (10 percent), and corruption (6 percent). Although forty-three cases (61 percent) had at least one dissenter, the other twenty-seven were unanimous decisions (39 percent).
Of the seventy cases, the Supreme Court decided almost 69 percent for and only 31 percent against the sitting government. Table A2 in the appendix summarizes case outcomes by type of case. The SC voted most often against the government in cases involving corruption (50 percent), elections (43 percent), and economic matters (38 percent). The majority of cases dealing with executive prerogatives (81 percent), bill of rights (69 percent), and separation of powers (68 percent) were decided for the government.
The percentage of antigovernment votes spiked in 1997, 2006, and 2010. In both 2006 and 2010, Macapagal Arroyo was the president (Table 4). Since 2015, antigovernment votes have noticeably declined. The three-year centered moving average (red line, Figure 2) has since declined further, with the SC almost consistently ruling for the Duterte administration.
Average Tenure of Justices by Administration
Note: For incumbent justices, the numbers are calculated assuming that they would serve until mandatory retirement.
Case Outcome by President, 1986–2019
Dynamics of Deciding Votes.
Note: Dots in panel (a) represent the proportion of pro-administration to total votes of justices (progov_ratio), calculated for each decision; dots in panel (b) show the proportion of decisions with dissenting votes per year (dissent). A dot takes a value of 1 if all the decisions made in a given year involved at least one dissenting vote, 0 if all decisions were unanimous. The horizontal axis is the year the case was decided. The red line shows the three-year centered moving average values over time.
Nevertheless, dissenting votes have remained high throughout the sample period, with peaks in 1995/1996 and 2011. Since 2015, despite a noticeable decline in antigovernment decisions, the dissent rate was still >0.6, indicating considerable disagreement among the SC justices (Figure 2).
High levels of dissent indicate another important factor: the role of “factions” of justices appointed by the same president. The SC under Duterte illustrates this dynamic well (Figure 2): justices appointed by the politically aligned administrations of Macapagal Arroyo and Duterte tend to vote together, and those appointed by the opposing political camp (e.g., Aquino III) often join the dissenters (Table 5). The pattern was particularly evident as the Duterte administration began, when appointees from all three administrations were present. It became less pronounced as retiring justices were replaced. By 2020, eleven justices had been appointed by Duterte, three by Aquino III, and only one by Arroyo.
Number of Coinciding Votes and Their Percentages, Duterte Administration (2016–20)
Note: The numbers in the left column and the first row represent judges in order of appointment. Judges numbered 1 to 9 in the table were appointed by President Arroyo, 10 to 15 by President Aquino, and 16 and above by President Duterte. Numbers in the left-lower triangle of this table represent “coinciding votes/total votes” by a pair of justices. Numbers in the right-upper triangle report the percentages of coinciding votes. There were no votes observed for the gray areas.
Individual Voting and Regression Findings
The voting records of individual justices reveal major differences in votes both for and against the government and in dissents. For instance, in our sample of eighty-six justices, four have not voted in megapolitical cases. Of the other eighty-two, forty-four have dissented from a majority opinion, and thirty-eight never dissented. That raises another question: do individual traits shape the voting patterns of SC justices? To find out, we engaged in basic regression analysis:
Our dependent variable (vote_for_government) is binary, with a value of one if the vote is for the administration, zero otherwise. Of the independent variables, those related to our hypotheses that relate to a strategic understanding of how justices might behave are:Footnote 15
-
• appointer: a dummy variable that that indicates if the justice is appointed by the current administration (H1)
-
• first-year_appointee: a dummy variable that represents a newly appointed justice who votes in a decision within a year of his or her appointment to the bench (H2)
We also added control variables that represent a justice’s characteristics and those possibly affecting a decision. Those include: gender, age, previous jobs, the law school from which the justice graduated, the type of cases, and under which presidential administration the case was decided. To capture the age effect, we constructed a variable (remaining_years_of_justice) that measures how much longer a justice was likely to serve before mandatory retirement when the decision was promulgated. Moreover, a variable (remaining_years_of_president) was added to control if decisions against the current administration are more likely toward the end of the single six-year presidential term, in line with possible strategic defection known from the literature (Helmke Reference Helmke2002).
The distribution of the 940 votes by the eighty-two justices in the seventy cases identified is far from balanced. The number of votes of individual justices ranged from 1 to 46, with the average being 11.5 votes, the median 9, and the mode 7. Therefore, a pooled cross-section Probit model was used, and the parameters are estimated by maximum likelihood. To account for the possible lack of independence among individuals voting on a case, observations were clustered by cases and the cluster robust standard errors of the estimated parameters were calculated.
Table 6 summarizes the marginal probabilities of important variables with various specifications. Model (1) is the benchmark model. In Models (2) to (5), dummy variables are added in order to control for various factors that may affect the decision. Since these control variables are dummy variables, rather than including everything at the same time, we added them to the benchmark model to see if the choice affects the results substantially. Although the reported marginal probability varies slightly by model, qualitative findings, as summarized below, are robust across models.Footnote 16
Marginal Probabilities Calculated from Probit Regression Results
Note: STATA17 is used for estimation. Additional dummy variables are included in Models (2) to (5). Since the votes on decisions by individual justices are likely to be dependent, cluster robust standard errors are reported in parentheses. ***, **, and * indicate the significance at 1%, 5%, and 10%. Count R 2 is the proportion of correctly classified predictions, by setting the threshold at 0.5.
Further, we investigate whether this voting behavior varies by presidential administration, giving special attention to variations of the appointer effect. We have extended the regression model so that the coefficient of appointer varies by president. Table 7 reports the corresponding marginal probabilities, representing a probability of pro-government ruling by justices newly appointed over incumbent justices, by administration.
Pro-Government Votes by Newly Appointed Justices by Administration
Note: Models in Table 6 are extended by including the interaction terms of the appointer and a set of president dummy variables. For the Corazon Aquino period (1987–92), the probabilities were not calculated since all the justices were newly appointed.
While the sample is small and the results should be interpreted cautiously, the tendency is robust across different specifications with different control variables for the following main findings:
-
1. The coefficient of appointer is positive. This implies that if the justice was appointed by the current president, this significantly increases the probability of a pro-government vote, as hypothesized in H1. The tendency is robust and about 15 percent higher.
-
2. The coefficient of first-year_appointee is also positive and statistically significant. The probability of pro-government votes is roughly 20 percent higher for freshmen, which is consistent with H2.
-
3. Finally, Duterte-appointed justices appear to have a probability of voting in favor of his administration that is higher by about 25 percent than the probability of justices appointed by Duterte’s predecessors voting in favor of the presidents that appointed them.
Combined, the results show that there are loyalty effects across administrations. These loyalty effects wax and wane depending on time on the bench, and they have been more pronounced under the Duterte administration.
CONCLUSION
The Supreme Court of the Philippines offers a fascinating but complex opportunity to study judicial behavior. One of the oldest courts in Asia, its powers and influence have waxed and waned throughout its 120-year history. During the Marcos dictatorship, it was widely perceived as having succumbed to the executive, but the return to democracy in 1987 saw an expansion of its powers. It has since emerged as one of the most recognized political institutions in the Philippines and one of the most activist courts in the region. And yet, as might be expected from an institution that has generally operated within a “low-quality” presidential democracy marked by elite dominance, institutional weakness, and widespread abuse of the highest office (Hutchcroft and Rocamora Reference Hutchcroft and Rocamora2003), the SC has also struggled to insulate itself against executive influence and elite infighting. As a result, it has been accused of partisan behavior in high-profile political cases—problems that have attracted renewed attention given the erosion of democracy since 2016.
Taking these public and academic concerns as a starting point, this article offers an extension of empirical scholarship on the behavior of SC justices in high-profile political cases, with particular attention to loyalty effects. Often legally ambiguous, megapolitical cases are particularly suited for investigation because it would be reasonable that strategic behavior, novelty of issues, as well as loyalty dynamics might come into play as they are being decided. While not a replacement for legal-interpretivist scholarship, our findings offer a nuanced and empirically grounded perspective on the SC track record over four decades and whether a new pattern of behavior has emerged under the populist-authoritarian rule of President Duterte.
With particular attention to possible loyalty effects, our study offers support for some claims. As H1 postulated, justices voting on cases resolved during the incumbency of the president who appointed them are more likely than those appointed by previous presidents to vote for the government in high-profile cases. The effect is statistically significant across all model specifications.
For H2, the “freshman effect,” we found that justices in the first year of their appointment are more likely to vote for the appointing president; among possible reasons are bewilderment, disorientation, or peer pressure. What precise dynamic is at play is hard to discern, but in a context like the Philippines, these could be reinforced by a culture of gratitude (utang na loob) and a sense of group conformity (pakikisama) previously described in the literature (Torres 1985, 494; Agabin Reference Agabin2012).
Finally, for H3, we found distinct differences in the behavior of justices in different administrations. In particular, findings in H1 are particularly pronounced under Duterte: the probability of Duterte appointees voting in favor of his administration is significantly higher than justices voting for the administrations of his predecessors who appointed them. Given his already numerous appointments, this finding may have major practical implications.
We note that under the Duterte administration, the SC continues to demonstrate judicial activism in high-profile cases. The number of megapolitical cases decided peaked in 2019 with a high number of disputes between branches of government. And while votes against the administration have been notably in decline under Duterte, dissents by SC justices have remained high. These suggest not only a degree of independence but also possible factional dynamics aligned with loyalties to appointing administrations.
There is much room to speculate why these patterns, both expected and unexpected, occur. What is clear is that traditional models of judicial behavior travel with difficulty to the Philippines: ideological affinities on the bench seem almost nonexistent and policy views are difficult to discern. Political parties and presidential elections are largely guided by personality preferences rather than programmatic policies.
Instead, our findings show that the Philippine SC draws attention to some less understood informal dynamics of presidential loyalties and related factional networks on the bench. Formed as part of small educational and professional pathways and reinforced by a culture or a debt of gratitude (utang na loob) and sense of group conformity (pakikisama), informal dynamics may shape judicial behavior in ways that strategic models do not account for. For instance, as suggested by our findings in high-profile cases, social network dynamics reinforce votes for the appointer and voting coalitions, and dissents mirror alliances between presidential administrations.
This is not to imply that the SC and its justices do not have agency. The Supreme Court of the Philippines is still an insulated institution with strong internal norms. As a strategic actor, it can protect its interests, especially when political actors have attempted to undermine it. And while the CJ has considerable internal power, individual justices have always appeared free to express themselves both on and off the bench. Nevertheless, despite greater public scrutiny and transparency over judicial nominations by the JBC, executive and third-party influence remain prevalent, as shown in the last decade by a number of scandals and the removal of two CJs.
The Duterte presidency is a good illustration of these dynamics and their consequences. Unlike most predecessors, he has appointed all but two members of the SC bench. Interesting new dynamics have emerged. A disproportionate number of new justices graduated from Duterte’s alma mater, San Beda Law School. Many of his appointments in the first three years were closer to retirement age, allowing him to appoint more justices to open slots at short intervals. By contrast, his appointees since 2019 have been younger, which extends his influence over the SC well beyond his presidential term.
The situation of the Philippine SC speaks to scholarly debates on the role of justices in fragile or fledgling democracies in Southeast Asia. Undoubtedly, many judges in the region struggle to deal with the tension between expectations derived from the law itself and persistent informal cultural demands and loyalties. This difficulty may well explain the unevenness sometimes exhibited in high-profile cases of international interest.
This study sheds light on the how loyalty effects may operate in less institutionalized settings than the US Supreme Court. It is particularly interesting to note that these dynamics are set in the unique context of a hyperpresidential system with a limited single six-year term. In the process, we hope that it will not only offer new comparative avenues, but also encourage a deeper contextualization of judicial politics in one of Asia’s oldest democracies.
Appendix
Descriptive Statistics of Core Variables
Outcomes by Type of Case
Note: The numbers in the second and third columns are the percentages of votes against and votes for the current sitting administration. The number in the rightmost column indicates the percentage of each case type.
Case List
Open access
