Shabbat and Shattered Dreams: Religious Accommodations for Public Exams in South Korea

Abstract

Today in South Korea, individuals of certain faiths are unable to take a wide range of state-administered qualifying examinations due to their religious convictions. The Constitutional Court of Korea has repeatedly refused their request for religious accommodations, such as an alternative test date for Sabbath or holy day observers who are unable to take exams on their original dates. The authors analyze the series of Constitutional Court decisions rejecting the need for such accommodation by focusing on the court’s use of its main analytical tool, the proportionality principle. These decisions reveal important shortcomings in the court’s application of the proportionality principle, including challenges inherent to proportionality and more specific deficiencies in the court’s application of the general principle. The article thus sheds light on how the proportionality principle is applied in the context of Korean constitutional jurisprudence and the resultant deprivation of protection for certain fundamental rights in Korea. The authors compare the court’s approach with that of courts in Spain, Switzerland, and the United States. They then propose a number of ways to improve the court’s proportionality analysis and its constitutional reasoning.

Introduction

In South Korea, public examinations—standardized examinations administered by or under the auspices of state agencies—serve as gateways to a wide range of careers, including in highly sought-after fields such as law, medicine, and civil service. For decades, the vast majority of public examinations have been held on Sunday or Saturday. ¹ There are administrative efficiencies associated with holding exams on weekends: more test venues are available during non-school days (many of the exams are held at school facilities) and proctors and test takers can avoid missing classes or taking time off from work to attend the exams. However, these weekend exams present a dilemma for test takers who observe a Sunday or Saturday Sabbath and whose religious convictions prohibit them from engaging in secular activities on the Sabbath.

Christians who observe the Sabbath or Lord’s Day are the group most affected by this issue. ² They must tread delicately when seeking religious accommodations for public exams. Making such requests in a fiercely competitive, traditionally collectivistic society has been difficult: Individuals who make such demands risk drawing the ire of both fellow test takers and the public at large, who are likely to denounce such actions as selfish and complain of unfairness. Accusations of foul play in entrance or certifying exams are extremely inflammatory in Korea, as illustrated by the fact that college admissions scandals contributed to both former president Park Geun-Hye’s ouster from power in 2017 ³ and the loss of former president Moon Jae-In’s Democratic Party in the 2022 presidential election. ⁴ Moreover, in a country where interest group politics are generally viewed with skepticism (professional lobbying is illegal in Korea, subject to limited exceptions ⁵), collective efforts by Christians to resolve the issue may well be castigated as political strong-arming by a cohort of religionists who have considerable political influence but nonetheless constitute only a minority of the population. ⁶ The situation is possibly worse for Saturday-observing Christians such as Seventh-day Adventists, who due to their smaller numbers have far less political clout to influence legislative or agency decisions that have a bearing on test scheduling, may (due to their putative “cult” status ⁷) experience difficulty collaborating with more mainstream denominations in their activism, and may even become inadvertent victims of efforts by other Christians to move exams away from Sunday.

There is currently little formal recognition of an individual’s right to seek religious accommodations for exams. Although legislative bills have sporadically been submitted to ban Sunday exams, none have been successful, and efforts to seek judicial remedies have likewise been mostly unfruitful. ⁸ There are practically no test administrators, public or private, that make religious accommodations available as a matter of policy. ⁹ Individuals must therefore seek accommodations on an ad hoc basis, often with a limited prospect of success.

Christian organizations have nonetheless persisted in their efforts to alleviate the situation for their members and such efforts have yielded some successes. For example, during the 2012 presidential election, a conservative Christian association succeeded in obtaining assurances from both major party presidential candidates that they would favorably consider holding public examinations on Saturday in order to avoid Sunday testing. ¹⁰ Although no formal change in the law or government policy ensued, there is some indication that state agencies have become more sensitive to the position of the Sunday-observing Christian community: For instance, all 37 public examinations administered by the Human Resources Development Service of Korea are held on Saturday (incidentally, the Sabbath observed by some Sabbatarians), the result of a gradual shift away from Sunday testing. ¹¹

On the judicial front, the Constitutional Court of Korea has heard a number of cases on the constitutionality of administrative decisions disallowing accommodations such as alternative test dates or make-up exams. The court—which is the final arbiter of all constitutional questions in Korea ¹²—ruled in favor of the public test administrator in all of these cases. The Constitutional Court’s stance on test accommodation stands in contrast to that of courts in a number of other jurisdictions with similar constitutional language protecting religious freedom. In the Korean cases, the Constitutional Court concluded that the claimed religious accommodations were not constitutionally required on the grounds that the administrative burdens and costs associated with providing such accommodations could not be justified. In other jurisdictions, including Spain, Switzerland, and the United States, courts have interpreted their respective constitutions to hold that religious accommodations for state-administered exams are in fact required as long as such accommodations are not excessively or unreasonably burdensome. ¹³

In what follows, we explore the reasons why, in Korea, the administrative burdens and costs for providing alternative test dates or sequestration continue to be recognized as sufficient justification to deny aspiring lawyers, teachers, and others from taking the necessary entrance exams—an issue that has largely escaped scholarly attention even in Korea. We analyze relevant decisions of the Korean Constitutional Court with a focus on the main analytical tool used to reach their conclusions, the proportionality test. The proportionality test as applied by the court in these cases suffers from important shortcomings that can reduce proportionality to a complete deference standard for governmental actions, especially in matters that lack popular support such as religious accommodation for exams.

Overview of Korean Constitutional Court’s Religious Freedom Jurisprudence

Freedom of religion is encapsulated in Article 20 of the Korean Constitution, which states that “[a]ll citizens shall enjoy freedom of religion” and that “[n]o state religion shall be recognized, and religion and state shall be separated.” ¹⁴ The freedom of religion, like any other fundamental right, ¹⁵ is subject to the limitation clause in Article 37, Paragraph 2, of the constitution, and hence subject to proportionality analysis.

The Constitutional Court has ruled that the constitutional freedom of religion consists of the freedom of faith, the freedom to engage in religious conduct, and the freedom of religious assembly. ¹⁶ Freedom to engage in religious conduct includes the freedom to act in accordance with religious convictions and doctrines, such as by voluntarily engaging in religious activities, including through participation in religious rituals and worship services; the right not to be coerced against one’s religious convictions; the freedom to proselytize; and the freedom of religious education. ¹⁷ The freedom of religious assembly is interpreted as the freedom of co-religionists to hold gatherings or to form religious organizations. ¹⁸ The freedom of religious conduct, unlike the freedom of faith, is not absolute, and may be subject to restrictions for purposes such as maintaining the order and public welfare. ¹⁹

In addition to the test accommodation cases, which we discuss in detail below, the Constitutional Court’s religious freedom cases generally fall into the following four categories: ²⁰

1. 1. Exemption of religious organizations from regulatory requirements. The court has issued a series of decisions in which it ruled that not exempting religious organizations from generally applicable regulatory requirements (such as an infrastructure surcharge on buildings that exceed 200 cubic meters, ²¹ registration requirements for schools and private educational institutions, ²² reporting requirements for eldercare facilities ²³) did not violate the freedom of religion. The court held that the constitution does not grant individuals the right to demand positive preferential measures regarding religion, ²⁴ and found that the regulations’ public interest benefits outweighed any infringement on religious liberties. ²⁵ Similarly, the court ruled that a law that accords tax exemption status only to religious corporations that meet certain requirements and not to religious corporations at large does not violate the freedom of religion. ²⁶

2. 2. Overseas travel to proselytize in high-risk areas. In a case where the complainants challenged a government restriction on traveling to Afghanistan and other countries with a high risk of terrorism on the basis that such measure prevented them from conducting missionary work in these countries, the court ruled that the restriction did not violate the freedom of religion. ²⁷

3. 3. The right of detainees/prisoners to attend religious events. The court has generally ruled that it was a violation of religious freedom for prisons to prohibit pretrial detainees from attending religious events. The court dismissed the prison’s argument that such a measure was necessary due to the danger that the detainees could use such gatherings to associate with other individuals, including co-conspirators, noting that there were alternatives available that were less restrictive of the detainees’ rights, such as having co-conspirators attend separate events. ²⁸ On the other hand, the court did not find a violation of the freedom of religion where a prisoner was held in a segregation unit for fifteen days and restricted from participating in communal events, including religious services during that time. ²⁹

4. 4. Conscientious objectors. For decades, the court maintained the constitutionality of statutory provisions penalizing conscientious objectors, most of whom were Jehovah’s Witnesses. ³⁰ In 2018, the court reversed its position and found such provisions to be a violation of the complainants’ freedom of conscience. ³¹

In general, the Constitutional Court’s religious liberties jurisprudence has not drawn much attention or been controversial. An important exception is the conscientious objectors ruling, which implicates significant harm to individuals on the one hand (imprisonment or criminal record) and important public interests (national security and fairness in conscription) on the other. The test accommodation cases, which implicate considerable harm to the relevant individuals in the form of a potentially permanent loss of career opportunities, have drawn little attention and evoked less debate.

Proportionality: Theory and Application by the Korean Constitutional Court

The Proportionality Test in Constitutional Review

For decades, the Korean Constitutional Court has applied the proportionality test to assess whether the exercise of public power and the resultant restriction of a constitutional right or freedom are “proportionate” in light of the justification provided by the legislature. The proportionality test dictates that an exercise of public power—such as a government measure—that significantly restricts a constitutional freedom would only be able to survive the proportionality test if there is, on balance, an equal or greater need for such a restriction. ³² For a less restrictive measure, the bar would be lowered accordingly. This balancing exercise, the essence of which is to weigh the harm to the constitutional right or freedom against the justification behind the rights-infringing measure, is observed in many different legal traditions, civil law and common law jurisdictions alike. ³³ Originally developed as a part of German constitutional law, this doctrine has spread widely to countries across the world, including South Korea, collecting titles such as “a central feature of rights adjudication in liberal democracies worldwide” ³⁴ and a “foundational element of global constitutionalism.” ³⁵ Proportionality, however, remains a controversial doctrine with many fierce critics who question its correctness and its effectiveness in constitutional review.

Rights, Interests, and Balancing

Rights advocates have long argued that proportionality downgrades the protection of fundamental rights to the level of policy arguments. ³⁶ Jürgen Habermas, one of the most well-known critics in this category, considers that proportionality deprives rights of their “strict priority” and makes them indistinguishable from interests, policies, or values. ³⁷ According to Habermas, in the proportionality world, rights are downgraded to values, and values are “inherently just as particular as every other” with no rational basis for priority. Value systems—as opposed to systems of norms or rights—are “flexible,” and can be traced back to “particular cultures” rather than being “fundamental” or “universally-binding.” ³⁸ Habermas is further worried that proportionality takes away the “correctness” or “legitimation” given to courts’ decisions, in light of the fact that values are a matter of preference. ³⁹

A related but different criticism of proportionality review centers on the lack of commensurability. The commensurability argument is different from the argument that constitutional rights and freedoms should categorically be exempt from any balancing because they should enjoy a certain priority. The premise is rather that even if one were to accept the balancing of rights against other interests, these competing rights, values, and interests are essentially incommensurable, that is, irreducible to a common measure. ⁴⁰ Without a common metric, the illusive balancing is simply impossible. These critics have pointed out that to engineer commensurability among the wide array of rights, one must “subscribe to some form of utilitarianism, namely, to a moral theory that assumes all interests are ultimately reducible to some shared metric (money or happiness or pleasure).” ⁴¹ In other words, under proportionality, rights will lose their normative priority and will become the subject of some type of utilitarian analysis that has no rational basis.

Some proponents have responded that proportionality—done properly—will conserve the normative priority of fundamental rights and be rational and structured. Alexy, for instance, provides a detailed schematic on how the proportionality test should be carried out by courts and argues that following such a structure, the prioritization of rights is possible. ⁴² Alexy argues that the proportionality test actually consists of three sub-principles as observed in the German constitutional law, with the problematic balancing and the potential for watering-down of fundamental rights really only appearing in the last of the three sub-principles. The three sub-principles are (1) the suitability test, which examines whether the means employed are suitable to achieve the statute’s intended purpose; (2) the principle of necessity, which examines whether the intended purpose can be achieved through less restrictive means; and (3) the principle of proportionality “as such” (proportionality stricto sensu), where the actual balancing happens. ⁴³

Alexy argues that if a statute or a measure fails either of the first two principles—suitability or necessity—the balancing of rights can simply be avoided because one will never reach the third sub-principle. Additionally, for this third sub-principle where the balancing occurs, Alexy suggests another set of three steps that allows one to measure the relative weights of the competing rights and interests: the first step concerns establishing the weight of argumentation regarding the intensity of interference with the right by using a scale of “light,” “moderate,” and “serious”; the second step concerns establishing the weight of argumentation regarding the intensity of satisfaction of the benefit achieved for the competing principle along the same scale as the first step; and the third step involves comparing the relative weights of the two. ⁴⁴ By pointing to a common scale that could be used to assess the competing rights and interests, Alexy argues that there is sufficient structure here to ensure that the proportionality analysis is neither arbitrary nor irrational. Even when balancing is called for, a rational method of comparison exists as long as courts faithfully engage in the argumentation concerning the intensity of the interference or satisfaction. ⁴⁵

Furthermore, Alexy argues that proportionality analysis—if properly done using the above-mentioned schema—will protect the normative priority of an important right. This is because a “center of resistance” exists for each right. The center of resistance denotes the core “area [of a right] where interferences can scarcely ever be justified by strengthening the reasons for the interference.” ⁴⁶ Alexy suggests that when there is an extraordinary interference with a right, the center or the core of such a right will resist balancing. Matthias Klatt and Moritz Meister further developed this idea and proposed a “soft trumping” approach to proportionality, where rights will be given priority over other interests “according to their weight, without assigning them a categorical priority.” ⁴⁷ Under the soft trumping approach, “fundamental” rights—such as the right to life—will be given greater weight than other rights or values; thus, while rights are not categorically exempt from balancing, they will prima facie trump lesser values or interests. ⁴⁸ Rights are therefore given “soft” priority and should be treated as “soft trumps.” Klatt and Meister also state that, under this approach, only those justifications that are of “constitutional status” deserve to be balanced against a fundamental right. ⁴⁹

In sum, proponents have responded to the criticism that proportionality is normatively problematic (downgrading fundamental rights) and irrational (requiring a utilitarian framework lacking any rational basis) by arguing that even in the world of balancing, if done properly, fundamental rights such as human dignity would be placed at a higher priority compared to other principles. They further contend that through employment of methodologies such as the three-principle approach outlined above, a rational and structured legal reasoning is possible. However, there is a large discrepancy between the theoretical account of proportionality done properly and how the tool is actually used in the public examination cases in Korea to protect the constitutional right to religious freedom.

Insufficient and Ambiguous Information

Critics have also voiced a second type of challenge based on the impracticability of proportionality as a tool for legal reasoning because it must rely on insufficient or ambiguous information. Critics have pointed out that it is very difficult, if not impossible, to gather sufficient information to properly apply (1) the suitability test, which assesses whether the means employed by the interfering statute or measure are suitable for achieving its aim; and (2) the necessity test, which assesses whether the aim can be achieved through any less restrictive means. Although these are empirical problems that should be decided based on empirical evidence, in an actual courtroom setting, “often all one has are assumptions, contradictory experiences, and as many expert opinions as there are interests involved.” ⁵⁰

Additionally, there is likely to be a significant information asymmetry between an individual whose fundamental right is restricted and the legislature or the government. Depending on burden-of-proof rules, individuals whose fundamental rights have been restricted may face the burden of showing that the rights-infringing statute or measure does not meet the necessity or the suitability test. The empirical information needed to do that is often inaccessible to the individual or too costly to obtain or analyze. Individuals often also lack the ability to adequately challenge and fight the competing versions of empirical information provided by the legislature or the government defending the measure.

Critics also point out that a statute or a measure rarely comes with a clear articulation of a single purpose. The text of the legislation and the surrounding legislative histories often allow for reasonable interpretations of multiple purposes. This is problematic for proportionality analysis, as how the purpose is delineated has a bearing on the outcome of (1) the suitability test, which assesses whether the means employed are suitable for achieving the intended purpose; (2) the necessity test, which assesses whether the intended purpose can be achieved through any less restrictive means; and (3) the proportionality stricto sensu test, where the intended purpose and the intensity of satisfaction are weighed against the degree of infringement on the individual’s rights. Mark Tushnet points out the difficulty in identifying a single legislative intent behind a statute enacted through a modern democratic legislative process. He argues that it is often the case that a myriad of permissive legislative purposes may be identified for a single piece of legislation. ⁵¹ The court lacks clear bases for articulating the “one” purpose of the statute in question. The court similarly lacks clear bases for selecting the “right” or “interest” at issue. Tushnet, who calls this process “conditionalizing,” points out that it is not possible to pursue this process non-arbitrarily. ⁵²

In response to such criticism, proponents have suggested well-designed burden-of-proof rules to solve the problem of insufficient or ambiguous information. Bernhard Schlink argues that many jurisdictions around the world have employed flexible burden-of-proof rules to overcome this problem. ⁵³ For instance, if the infringed right is a fundamental right of an individual and the rights-infringing statute’s purpose is relatively minor, courts may require the legislature to bear the burden of proof concerning suitability, necessity, and balancing. ⁵⁴ Under such a scenario, the rights-infringing statute would be presumably unconstitutional and the individual would not have to carry the burden of proving that the statute is not suitable, not necessary, or disproportionate.

Proponents have similarly responded on whether it is possible to identify the single intent behind a single piece of legislation. Aharon Barak writes: “members of the legislative body, collectively, devise a purpose. If they reach an agreement and if they constitute a majority, they would vote and enact a law which, in their opinion, is aimed at achieving that purpose. … The refusal to accept the notion of legislative intent is much like a refusal to accept the notion of legislation at all.” ⁵⁵

In sum, the proponents argue that the problem of insufficient or ambiguous information may be overcome by well-designed burden-of-proof rules and careful review of the intent of the legislative body. In all of these responses by proportionality proponents, the theory of proportionality—the ideal—is presented to dispel the criticisms, which is hardly observed in the actual utilization of the tool in the public examination cases in Korea.

The Proportionality Test as Applied in Korea

Constitutional Basis of the Court’s Proportionality Test

Throughout its thirty-some year history, the Korean Constitutional Court has relied primarily on the proportionality principle to justify its reasoning in constitutional jurisprudence. ⁵⁶ The Constitutional Court’s use of the proportionality test has been justified under Article 37, Paragraph 2, of the constitution, the so-called limitation clause: “All freedoms and rights of citizens may be restricted by statute only when necessary for national security, the maintenance of order or for public welfare. Even when such restriction is imposed, no essential aspect of the freedom or right shall be violated.” ⁵⁷

The court has interpreted “only when necessary” to signify “only to the extent necessary” so that a restriction to a citizen’s freedom or right must be not greater than necessary—that is, it must be proportional—to achieve the purposes of national security, maintenance of order, or public welfare. ⁵⁸

The limitation clause refers to “[a]ll freedoms and rights,” which suggests that no right or freedom is completely sacrosanct and all may be subject to some limitation. ⁵⁹ At the same time, the “no essential aspect of the freedom or right shall be violated” language suggests a version of the rights as trumps model advocated by Habermas and others, at least for some “essential” features of the restricted rights. The Constitutional Court has given relatively little attention to the prohibition on infringing an “essential aspect,” however. ⁶⁰

Elements of the Korean Constitutional Court’s Proportionality Test

Heavily influenced by German constitutional law, Korean constitutional jurisprudence often draws upon the work of German theorists and courts when analyzing constitutional concepts, including the principle of proportionality. ⁶¹ As articulated by the Constitutional Court, the four elements of the proportionality rule are (1) “justifiability of purpose” plus the three elements that are also considered by the German courts; these are (2) suitability (Geeignetheit); (3) necessity (Erforderlichkeit), also referred to as minimal impairment; and (4) proportionality in the narrow sense, also known as proportionality stricto sensu (Verhältnismäßigkeit im engeren Sinne).

According to the court, the “justifiability of purpose” step examines whether “the purpose of a law that would restrict a fundamental right of the citizenry must be deemed justifiable in light of the constitutional and statutory system.” ⁶² The court has interpreted this element to require that the relevant restriction must serve one of the three broad public interest goals enumerated in Article 37, Paragraph 2, of the constitution: national security, maintenance of order, and public welfare. ⁶³ In the majority of proportionality cases, the court’s discussion on this prong remains minimal and the court generally finds that the purpose of a restrictive statute or measure is in fact to protect the relevant public interest goal. ⁶⁴ Next, the “suitability” element examines whether the law is effective and appropriate as a means for achieving the intended purpose. ⁶⁵ Overall, the court tends to defer to the legislature’s choice of means for achieving a particular outcome. ⁶⁶

In the “minimal impairment” (or “necessity”) step, the court examines whether the right or freedom is infringed to the minimum extent necessary. ⁶⁷ To satisfy this step, the measure taken must be the least restrictive alternative. ⁶⁸ In general, the court spends the most effort discussing minimal impairment while often glossing over the other sub-tests. ⁶⁹

Lastly, in the proportionality stricto sensu step, the court examines whether the public interest to be protected is, on balance, greater than or equal to the private interest that is restricted. ⁷⁰ Scholars of Korean constitutional law generally agree that proportionality in the narrow sense is where the actual balancing occurs and hence is the most important part of the test. ⁷¹ In the court’s actual practice, however, proportionality stricto sensu is usually subsumed into the necessity analysis and not reviewed as a truly independent prong. ⁷²

Despite the court’s accumulated jurisprudence, the court’s analyses and application of the elements of proportionality have remained rudimentary; in the words of one scholar, the court has employed superficial analysis “that hurries toward a hastily determined conclusion.” ⁷³

The Korean Constitutional Court’s Jurisprudence on Religious Accommodation for Exams

The Korean Constitutional Court’s “test accommodation” cases concern whether public institutions must accommodate potential test takers who cannot take a government-administered exam on the scheduled date due to religious convictions. The court has ruled on this issue six times: in 2001, regarding the National Judicial Examination (2001 Judicial Exam Case); ⁷⁴ in 2010, again on the National Judicial Examination (2010 Judicial Exam Case); ⁷⁵ in 2010, regarding the Legal Education Eligibility Test (2010 LEET Case); ⁷⁶ in 2010, regarding the public school teacher appointment examination (2010 Teachers Exam Case); ⁷⁷ in 2022, on the Bachelor’s Degree Examination for Self-Education (2022 BDES Case); ⁷⁸ and in 2023, on the assistant nurses examination (2023 AN Exam Case). ⁷⁹ A review of these cases highlights the fundamental challenges to the court’s use of the proportionality analysis in this context.

2001 Judicial Exam Case

This case concerns the National Judicial Examination, the old form of Korea’s bar exam. ⁸⁰ The National Judicial Exam had to be taken by anyone who wished to enter the legal profession. It was a highly competitive, high stakes exam with a passage rate in the single digits that was historically seen as an important opportunity for those with less means to climb the social and economic ladder through hard work. ⁸¹

The complainants in this case were aspiring members of the legal profession who happened to be devout Christians. They were unable to sit for a National Judicial Examination scheduled for Sunday and brought a constitutional complaint against the government agency responsible for administering the exam. The complainants claimed that the order scheduling the exam for a Sunday ⁸² amounted to religious discrimination and infringed their religious freedom. ⁸³

In a brief opinion, the court disagreed and denied the complainant’s claim that the Sunday-only examination schedule is an unconstitutional restraint of the complainant’s right to religious freedom. The court began by identifying the restricted right as the freedom to manifest religious beliefs, which is distinct from the freedom to hold religious beliefs. The court ruled that the right to manifest religion is not an absolute right and may be restricted for purposes such as maintaining order or public welfare. ⁸⁴ The court then proceeded to conduct an extremely abbreviated proportionality analysis, which neither clearly distinguished each of the sub-elements nor meticulously followed the logic articulated for each sub-element in its other opinions.

First, the court implied—without much evidence or discussion—that the purpose behind the government’s order of scheduling the examination on a Sunday is to maximize the convenience of the majority of test-takers. The court also implied that such purpose is appropriate by interpreting a provision of the State Public Officials Act, a statute that governs matters relating to public servants. ⁸⁵ Article 35 of the Act reads, “Any examination for appointment through open competition shall be open equally to all citizens having equal qualification, and the date and venue of the examination shall be decided by taking into consideration the convenience of the examinees.” ⁸⁶ Without further analysis, the court agreed with the government’s position that “schedules for public exams should be set so that it imposes the least burden on the everyday livelihood of the majority of citizens and not on a particular religion.” ⁸⁷ The court’s analysis of the justifiability of purpose element was conducted with little to no explicit discussion of the order itself, the governing statute or their legislative histories.

Second, with the purpose of the order identified as the convenience of the majority, the court repeated the justifications advanced by the government for scheduling the National Judicial Examination on a Sunday to conduct its suitability and necessity analyses: the large number of public officials needed in order to manage an examination of that scale and difficulties attendant to holding the exam on a weekday (disruption to school schedules at middle and high school facilities used as test venues, inconvenience to test takers who were students or professionals, and disorder caused by question booklets being carried in at around the same time examinees arrived at the testing location). The court implied—without any further discussion—that for these reasons, the Sunday-only scheduling order was a suitable and necessary method to achieve the purpose of protecting the convenience of the majority. ⁸⁸ The court then concluded its proportionality analysis by declaring that “the extent of [the restriction on the complainants’ right to religious freedom] cannot be regarded as overstepping the principle of proportionality [stricto sensu], nor as an infringement on a fundamental aspect of the complainant’s religious liberty.” ⁸⁹

2010 Judicial Exam Case

Although the court had rejected the complainants’ plea for the National Judicial Examination to be held on a day other than Sunday in the 2001 Judicial Exam Case, the government in 2010 scheduled the written component for the exam for a Saturday, instead of a Sunday. Some speculated that it was the government’s experience of defending the 2001 Judicial Exam Case and the presence of a large Sunday-observing Christian community in Korea that motivated the change in exam dates to a Saturday. ⁹⁰ While the change might have placated Sunday-observing Christians, it had a detrimental effect on a different religious group. In the 2010 case, the complainants were Seventh-day Adventists, who observe a Saturday Sabbath.

While the surrounding facts and the arguments were largely similar to those in the 2001 Judicial Exam Case, there were also significant differences: the exam at issue was held on Saturday rather than on Sunday, the complainants were members of a minority Christian denomination, and the complainants challenged not only the constitutionality of the scheduling order itself but also the government’s denial of their request to be sequestered until sundown so that they could start taking their exam after Sabbath.

Despite these differences, the court’s substantive analysis remained almost identical to that in the 2001 Judicial Exam Case. The court repeated the highly abbreviated proportionality analysis concerning the complainants’ right to religious freedom. The court stated that the freedom to manifest a religion may be limited in order to maintain the order or for public welfare. It cited the same Article 35 of the State Public Officials’ Act provision to identify the purpose of the scheduling order and governing laws as promotion of the convenience of multiple test takers, and held that the justifiability of purpose prong was satisfied. The court also found that scheduling the test on a Saturday was a suitable and necessary method to achieve such a purpose, observing that it is “desirable” to set the test date to minimize the number of those negatively impacted and citing similar difficulties with holding the examination on a weekday as noted in the 2001 case. ⁹¹ The court further examined the option of allowing the complainants to test separately after sundown on Saturday as a less restrictive alternative but held that such arrangement entailed the risk of cheating and complaints from other test takers and “various difficulties in test management.” ⁹² Moving on to the proportionality stricto sensu test, the court stated that whereas it is difficult to conclude that administering the test on Saturday directly infringed on the complainant’s right to manifest religious beliefs or a fundamental aspect of the freedom of religion, the decision to administer the exam on Saturday made a substantial contribution to the convenience of test takers, procurement of testing locations and convenience of test management. ⁹³

2010 LEET Case / 2010 Teachers Exam Case

The complainants in both these cases were Sunday-observing Christians, who challenged the constitutionality of the LEET Exam (law school entrance exam) and Teachers Exam (public school teacher appointment examination) that were only administered on Sunday. ⁹⁴ The court cited to the two National Judicial Examination cases and repeated its arguments from those cases. ⁹⁵ Notably, the court did not address the fact that a national examination of a comparable scale (that is, the National Judicial Examination) had been successfully held on Saturday that year, which arguably weakened the government’s arguments concerning the impracticality of rescheduling the exam to a Saturday as requested by the complainants. The court also did not note a fact that it could easily have cited as a justification for its decision—that granting the complainants’ request for holding the test on Saturday could result in an infringement of the right of test takers who happened to be Saturday Sabbatarians, like the complainants in the 2010 Judicial Exam Case.

2022 BDES Case

Twelve years later, the court issued a ruling in a constitutional complaint concerning scheduling decisions for the Bachelor’s Degree Examination for Self-Education. The BDES is an examination that allows individuals who have not attended college or university to obtain a bachelor’s degree. There are four stages (referred to as courses) of the BDES, all of which an individual must pass in order to qualify for a bachelor’s degree. ⁹⁶ The complaint at issue was brought by a Sunday-observing Christian who challenged the test administrator’s decision to hold the test for all four courses in 2021 on Sunday. The court ruled against the complainant based on reasoning that is similar to that in the earlier test accommodation cases. ⁹⁷

The court articulated the purposes of the scheduling decision as ensuring maximum accessibility to test takers by minimizing interference with their daily obligations, including academic and work responsibilities; minimizing the burden to the relevant test facility; and facilitating the procurement of testing locations and test management. The court found the relevant scheduling decision to be a suitable means of achieving such purposes. Regarding the proposed alternative of holding the BDES tests during the week or on Saturday, as commonly done for other public exams, the court focused on the composition of the test taker population to support its conclusion that testing on Sunday was preferable to the other options. Specifically, it referenced statistical evidence from 2021, which showed that test takers encompassed not only students (33.8 percent) but also workers (19.2 percent), military service members (4.5 percent), self-employed individuals (2.6 percent), employees of public institutions (2.4 percent), and nurses (2.4 percent). Based on such data, the court emphasized the importance of considering the likelihood that many individuals would be required to work during weekdays and would possibly also have work commitments on Saturdays, as “many self-employed businesses work on both weekdays and Saturdays.” ⁹⁸ The court did not address empirical evidence on the extent to which self-employed businesses also open on Sundays.

Finally, in the proportionality stricto sensu analysis, the court noted the complainant’s ability to attend church services outside the designated test hours and deemed the extent of infringement to be relatively minor. Weighing such minimal infringement against the costs and administrative burdens associated with holding the exams during the week or on Sunday, the court concluded that the measure successfully satisfied the proportionality test. ⁹⁹

2023 AN Exam Case

The complainant in this matter was a Seventh-day Adventist who had completed the necessary theoretical and practical training to qualify for the assistant nurses’ certification exam. The examination was held twice a year, both times on Saturday. The relief sought was for the test administrator to either diversify the test dates so that at least one of the exams each year would be held on a day other than Saturday or allow the complainant to sit for the exam after sundown on Saturday. ¹⁰⁰

In a 6–3 ruling, the court dismissed the complaint. The court’s proportionality analysis again closely mirrored the precedents. The court accepted the government’s arguments that it would be costly and inefficient to hold the examination on a weekday. It dismissed Sunday testing as a reasonable alternative, observing that such a course would infringe the right of Sunday-observing Christians and thus fail to resolve the issue of religious liberties infringement. The court did not address the possibility that holding the exam alternately on Sundays and Saturdays could provide both Saturday and Sunday observing Christians with the opportunity to take the exam without compromising their religious beliefs. The court also found that it would be administratively burdensome, and possibly unfair, to allow the complainant to take the test after sundown on Saturday. ¹⁰¹

For the first time in the test accommodation cases, a minority opinion was issued by the dissenting judges. The dissent agreed that the test scheduling decision satisfied the justifiability of purpose and suitability prongs, but found that it failed the necessity and proportionality stricto sensu tests. The dissent opined that holding both exams each year on Saturday was not the least restrictive option for achieving the purposes of the scheduling decision, which the majority had described as promoting the convenience of test takers, minimizing the burden to test facilities, and facilitating venue procurement and test management. The dissent suggested that holding the exams alternatively on Sunday and Saturday, or conducting the exams on Saturday while allowing the complainant to take a make-up exam on other days (as was the practice in certain countries), would have been less restrictive alternatives. The dissent noted that Seventh-day Adventists would not be able to take the exam without violating their religious convictions if all the exams were held on Saturday and portrayed the test scheduling decision as requiring the Seventh-day Adventist test takers to sacrifice for the sake of the majority’s convenience. The dissent found the harm to the Adventist test takers to be significant, and concluded that the administrative justifications put forth by the government were insufficient to outweigh such substantial interests. ¹⁰²

Summary

In six separate cases brought over the span of more than two decades, individuals who had been prevented by religious convictions from participating in various qualifying exams sought redress before the court, to no avail. The court’s reasoning in all of these cases was strikingly similar and showed little development over the course of these cases. Essentially, the court found in each case that scheduling the test for a Saturday or Sunday (as applicable) was suitable, necessary, and proportionate to achieve the purpose of maximizing test taker and administrative convenience, based on an abbreviated proportionality analysis that lacks methodological soundness.

Analysis of the Korean Constitutional Court’s Proportionality Assessment in Test Accommodation Cases

As discussed above, proportionality in general and the court’s application of the principle have been criticized on numerous grounds. The test accommodation cases exemplify various theoretical and empirical difficulties in the court’s proportionality analysis, including (1) lack of prioritization of fundamental rights; (2) failure to adequately assess the intensity of interference on fundamental rights; (3) de facto resort to utilitarianism and a resultant majoritarian bias; (4) insufficient inquiry and review of factual evidence and de facto placement of the burden of proof on the complainants; (5) difficulties in identifying the purpose of a statute or government measure; and (6) scrambling of the minimal impairment and proportionality stricto sensu steps.

Lack of Priority for Fundamental Rights

The religious accommodation cases exemplify how proportionality can contribute to the de-prioritization of fundamental rights. As in other religious liberties cases, the court has drawn upon the distinction between the freedoms to hold religious beliefs and to manifest religious beliefs, holding that the latter is not absolute and may be subject to restrictions. ¹⁰³ As noted in “Statutory Basis of Korean Constitutional Court’s Proportionality Test,” Article 37, Paragraph 2, of the constitution holds that “all freedoms and rights of citizens”—not just a subset of them—may be restricted when (and only when) “necessary for national security, the maintenance of order or for public welfare.” ¹⁰⁴ Thus, the court’s statement on its face does not render freedom of religious manifestation less constitutionally protected than other fundamental rights—all are subject to the limitation clause of Article 37, Paragraph 2, and the limitation clause itself serves both to allow restrictions based on certain enumerated public interests and to limit such restrictions to only the extent necessary. Nonetheless, the court takes this simple observation, essentially that the freedom to manifest religious beliefs is subject to the limitation clause, and uses it to justify treating the right to manifest a religion as a negligeable interest that can be overridden based on an assertion of administrative difficulties, whether or not robustly substantiated.

It may be meaningful to distinguish between freedoms concerning religious beliefs and the manifestation of such beliefs, as the latter is more likely than purely internal convictions to clash with or impinge on the rights of others or the public interest. Such distinction and the idea that the right of manifestation can be subject to limitations can also be found in international law: The International Covenant on Civil and Political Rights, for example, provides that “[n]o one shall be subject to coercion which would impair his freedom to have or to adopt a religion or belief of his choice” and that “[f]reedom to manifest one’s religion or beliefs may be subject only to such limitations as are prescribed by law and are necessary to protect public safety, order, health, or morals or the fundamental rights and freedoms of others.” ¹⁰⁵

However, these two aspects of religious liberty are not as distinct as the court’s decisions suggest: Freedom of manifestation is the right to act in accordance with one’s sincerely held religious beliefs and is thus not solely in the realm of external action; conversely, it is doubtful that the freedom to have or adopt a religion would be meaningful in a situation where the external practice of key tenets of that religion are prohibited or prohibitively difficult. Thus, it is reductive for the court to merely characterize a particular religious freedom case as a belief case or a manifestation case; such an oversimplistic classification scheme, coupled with the assumption that the right to manifestation is generally less worthy of constitutional protection, can lead the court to easily dismiss complaints in the so-called manifestation cases even in matters where the infringement on the individual’s religious freedom may actually be quite significant.

Fundamentally, the court’s test accommodation jurisprudence displays a failure to recognize the general priority of fundamental rights as such. The court’s approach contrasts with that taken by the Supreme Court of Spain in a test accommodation case brought by a Seventh-day Adventist who had been denied her request to take a state-administered language exam on a day other than Saturday. There, the court, interpreting the relevant statute, ¹⁰⁶ observed that “the doctrine of the Constitutional Court has repeatedly affirmed the greater value of fundamental rights and called for interpreting the legal system in the manner most favorable to their effectiveness” and granted the plaintiff’s request for a make-up exam. ¹⁰⁷

The Korean Constitutional Court, on the other hand, has dismissed complaints in the test accommodation cases without probing deeper into the value of the fundamental right at issue. In these cases, the court thereby validates critiques that proportionality is an inadequate mechanism for protecting fundamental rights over other interests. ¹⁰⁸

Inadequate Assessment of Degree of Interference

The court’s test accommodation jurisprudence shows the practical difficulty of realizing Alexy’s ideal for conducting proportionality analysis—categorizing the seriousness of the interference as well as the seriousness of the public interest justifications—particularly in connection with the court’s failure to assess in any depth the intensity of interference. In these cases, the court fails to analyze factors pertinent to the degree of interference with the complainants’ right to religious freedom, such as the significance of the practice of observation of the Sabbath and its relationship to the complainants’ religious beliefs, the consequences for the complainants individually and for adherents of the relevant religion collectively, and the ramifications for the society at large when certain constituents are restricted from pursuing various careers due to their religious beliefs.

Instead, the court simply noted that the complainants suffered a restriction on their right to manifest their religion. ¹⁰⁹ Omitting to further elaborate on relevant religious, individual, and societal implications allowed the court to discount the significance of the infringement, as shown for example in its simple restatement of the complainants’ hardship as “practical difficulties or disadvantages in [their] participating in worship service, etc.” ¹¹⁰ The court’s reasoning is unfortunately far from the analytical structure advocated by Alexy. Notwithstanding arguments by advocates that proportionality promotes a culture of justification that confers legitimacy on decisions that inevitably require consideration and weighing of various competing interests, ¹¹¹ in practice, the court’s proportionality analysis has failed to rise to the task of adequately constraining judicial discretion and ensuring protection for fundamental rights.

Incommensurability of Different Rights and Values, Utilitarianism and Majoritarianism

In the test accommodation cases, the court assigns less weight to the complainants’ religious freedom than to the administrative considerations advanced by the government, without enunciating why it chose to apportion the respective weights in this manner. This is partially attributable to the court’s slapdash reasoning and is likely also a byproduct of the difficulties inherent to identifying an objective metric for determining the relative weights of the competing interests. Ironically, such difficulties serve to lend a veneer of legitimacy to the utilitarian approach that is adopted by the court; utilitarianism, at the least, provides a putatively rational framework for conducting the balancing exercise.

As noted, the court in the National Judicial Examination cases described the purpose of the government’s scheduling decisions as promoting the convenience of the majority of test takers, based on a loose reading of Article 35 of the State Public Officials Act. Although the court did not articulate its reasons for delineating the purpose as such, a plausible reading is that this utilitarian construction was employed as it supports a similarly utilitarian approach in assigning the respective weights of the interests at issue. Once the grounds have been established for applying a utilitarian lens, the court’s task is simplified—the court can then easily find that tangible administrative costs and potential inconvenience to a majority of test takers outweigh the harms to the much smaller number of persons who are unable to test because of religious convictions.

The court’s utilitarianism in these cases contributes, naturally, to majoritarianism. The effects of such majoritarianism are the most pronounced in the 2010 Judicial Exam Case and the 2023 AN Exam Case, which were brought by complainants who were Seventh-day Adventists, a minority Christian denomination that mainstream denominations in Korea often denounce as a cult. ¹¹² An argument could be made that the court should have been even more sensitive to the potential impact on minorities when the case concerned a minority religious group stigmatized as a cult.

In contrast, in a test accommodation case before a Swiss public court brought by a Seventh-day Adventist plaintiff who had requested an accommodation regarding baccalaureate exams scheduled for Saturday, the court noted as follows: “Since school calendars generally already take into account the holidays celebrated by the traditional and most widespread religions, case law has established that the needs of individual religious communities, and therefore particularly of minority communities, must be taken into account insofar as they are still compatible with the public interest in ensuring orderly and efficient schooling.” ¹¹³ A similarly explicit consideration of the needs of minorities would help mitigate the utilitarian and majoritarian tendencies of the Korean Constitutional Court’s proportionality jurisprudence.

Insufficient Inquiry and Review of Evidence and De Facto Burden on Plaintiffs

The test accommodation decisions also evince a failure to carefully examine the validity of empirical arguments and evidence submitted by the parties. This problem is illustrated in the three test accommodation cases that the Korean Constitutional Court considered in 2010 and the more recent 2022 BDES Case.

In the 2010 LEET Case and 2010 Teachers Exam Case, the complainants argued that scheduling the tests for a Saturday would be a feasible alternative to holding the tests on Sunday and that the tests should therefore have been held on Saturday. ¹¹⁴ The government submitted various arguments on why holding the tests on Saturday would present difficulties, and the court held in both cases that holding the tests on Saturday was not an appropriate alternative. ¹¹⁵

However, in the 2010 Judicial Exam Case, the exam at issue (which involved a comparable number of test takers as the other two 2010 cases ¹¹⁶) was held on a Saturday. The court accepted the government’s arguments without questioning the evidence offered or inquiring why—despite the administrative difficulties asserted in the other cases—the government had chosen to hold the test on Saturday and how it had been able to successfully overcome such difficulties. In effect, over the course of three decisions on the exact same legal issue that were rendered over the course of seven months, the court alternated between agreeing that Saturday was not suitable for administering large-scale national exams (2010 Teachers Exam Case, 2010 LEET Case) and endorsing the holding of a particular large-scale national exam on a Saturday (2010 Judicial Exam Case), without even bothering to distinguish between the two lines of cases. Similarly, in the 2022 BDES Test, the court accepted the public test administrator’s argument that holding the exam on Sunday was preferable over Saturday testing due to the significant number of workers and self-employed individuals among the test taker population, without even pointing to concrete data on the extent to which businesses open on Sundays or analyzing differences with other exams that were held on Saturday.

In addition, while the court engaged in a superficial discussion of some numerical evidence in its opinions, the court failed to seek or review pertinent pieces of evidence such as the actual or estimated costs that would be associated with sequestering the complainants for a few hours or providing an alternative or an additional exam, and the potential damage or harm to the complainants or to society at large due to the exams being offered only once a year on a day that coincides with the complainants’ holy day.

The court’s inadequate examination of the evidence, coupled with its utilitarian approach in assessing relevant rights and interests, results in the complainants’ bearing the de facto burden of proving that the government’s measure fails the proportionality test. That is, although the decisions are silent on burden of proof allocation, in practice, all the government has had to do is adduce prima facie evidence of administrative costs and the burden then appears to shift to the complainants to show that the harms they suffer outweigh such costs. Proving this is inherently difficult given the more abstract nature of the harm to the complainants, ¹¹⁷ and the difficulty is compounded by the information inequality between the government and the complainants, as the complainants are often not in an optimal position to produce empirical evidence refuting the cost-based evidence submitted by the state.

The court’s perfunctory examination of the evidence and placing the burden on the complainants contrasts with the approach of courts in other jurisdictions in their test accommodation rulings. In the above Spanish Supreme Court case, the court held that “the cause that prevents the test or examination from being held on an alternative date must be of sufficient importance and … the Administration must highlight it with precision,” and ruled that the government had failed to meet this burden. ¹¹⁸

The Swiss public court also closely scrutinized the government’s argument that holding make-up exams on a different day based on a different set of questions would require “excessive administrative and preparatory effort … calling into question the orderly and efficient functioning of the school.” The court rejected such argument considering the fact that make-up exams were already allowed for illness, accident, bereavement, and coincidence with other exam dates; the small number of Seventh-day Adventists test-takers, which meant that only a few were likely to avail themselves of such an accommodation; the feasibility of obtaining reliable and objective results through a make-up examination; and the practice of other European Union Member States such as Italy and Germany. ¹¹⁹

In the United States, the Seventh Circuit reversed and remanded a summary judgment against a Jewish plaintiff who had been unable to take a civil service exam that was held on Saturday. The court found the government’s argument on the excessive costs of administering a make-up exam insufficiently substantiated, noting, “the plaintiff’s evidence of the personnel practices of other units of government which we have discussed above tends to indicate that these costs are not unreasonable or unduly burdensome … The stipulation of facts states that it would cost $1,600 to pay a personnel expert to design a comparable exam, pay for clerical help, and hire two proctors to administer a separate exam. This figure, however, does not relate to the sequestrations alternative suggested by the plaintiff.” ¹²⁰ The court also ruled that the government had failed to substantiate its arguments on the threat of litigation it would face if the accommodation was provided, because “there is no indication in the record that prior litigation involved circumstances similar to this case,” and that “our examination of Illinois law above convinces us that any litigation, if commenced, would be unlikely to be disruptive,” and “[l]itigation challenging agency action is almost always a possibility and therefore is a matter that the agency’s lawyers will naturally consider.” ¹²¹ These cases illustrate that with or without proportionality, ¹²² the judiciary’s careful scrutinization of relevant evidence can help courts to better assess the degree of burden that would accompany granting the requested protection for fundamental rights.

Conditionalizing: Framing of Statute’s Purpose

In the national judicial exam cases, the court refers to Article 35 of the Public Officials Act to find that the purpose of the statute governing the administration of public examinations is to ensure the convenience of the majority of test takers. ¹²³ In doing so, the court disregards the first part of Article 35’s language. Article 35 essentially consists of two parts: first, it states that open competition public exams should be “open equally to all citizens having equal qualification.” The second part of the provision reads, “the time and place of the examination shall be decided taking into consideration the convenience of the applicants thereof.” ¹²⁴ The first part—the principle of “equal access to all equally qualified”—seems to have escaped the court’s attention. The court appears to rely solely on the “convenience” language of the second part to articulate the purpose of the statute. If the court had given greater weight to the equal access principle when framing the purpose of the Public Officials Act, the results of the court’s proportionality analysis could have been fundamentally different. In other words, if the purpose of the provision is articulated so that greater emphasis is given to the requirement that equally qualified citizens should all be able to access the examination, the existing procedures that effectively prevent certain religious groups—regardless of their qualifications—from sitting in the exam may not be deemed suitable to achieve such purpose.

Instead, the court framed the purpose to be the promotion of test takers’ convenience and the facilitation of exam management. Furthermore, in the 2001 Judicial Exam Case, the court explicitly articulated that the purpose is to further the convenience of the majority of test takers. ¹²⁵ Once the purpose had been framed in this manner, the suitability and the necessity analyses were conducted in light of only such purpose and without consideration of the equal access principle. The court’s selective framing of the statute’s purpose lacks any evidentiary support.

Conflation of Minimal Impairment and Proportionality Stricto Sensu

The court has been criticized for scrambling the necessity (minimal impairment) and proportionality stricto sensu steps of the proportionality analysis. For example, Justice Jin-Seong Lee noted in a concurrence to a 2017 case that “despite the Constitutional Court’s thirty-year history, it is difficult to find any precedents where the necessity principle was met but the proportionality stricto sensu principle was not. This is because the assessment with respect to proportionality stricto sensu is conducted in the necessity section.” ¹²⁶ According to Justice Lee, strictly distinguishing between the necessity and proportionality stricto sensu analyses could, among others, enhance the logical nature and consistency of judgments through a more thorough examination of the importance of the public interest and the restriction to the individual’s fundamental rights.

In the test accommodation cases, the court likewise fails to flesh out its proportionality stricto sensu analysis. The negative effects of this are illustrated in the 2010 Judicial Exam Case. In its decision, the court allocates three paragraphs in total to the “necessity” and “proportionality stricto sensu” steps of the proportionality analysis. ¹²⁷ Of these three paragraphs, the first is devoted to discussing the administrative difficulties that would attend holding the test on a weekday, while the second paragraph deals with accusations of unfairness as well as administrative burdens that could result if the plaintiffs were allowed to take the test after Sabbath hours. In the third and shortest paragraph, the court indicated in one sentence that the government’s measure passed the proportionality test because there was no direct infringement of the complainants’ freedom to manifest religious beliefs or of an essential aspect of such rights, whereas the measure contributed greatly to the public interest (test takers’ and test administrators’ convenience). Essentially, the court spent the bulk of this section discussing the costs and inefficiencies attendant to the alternatives proposed by the complainants and then abruptly reached a conclusion on the proportionality stricto sensu prong based on the exact same considerations as outlined under the necessity prong plus a conclusory statement that there was no direct or serious infringement on the complainants’ right to manifest their religious belief. In other words, the court skipped a logical step in the proportionality stricto sensu analysis but this omission went undetected before the decision went to print, likely due to the court’s traditionally heavy reliance on the necessity prong.

Summary

It may be argued that in the test accommodation cases the Korean Constitutional Court essentially turned the proportionality test into something akin to a complete deference standard. Although the complete range of evidence and arguments considered by the court is not in the public record, ¹²⁸ the decisions themselves suggest that all the government had to do to prevail in these cases was to assert some administrative burden and that the court did not bother to inquire further on the actual merit of such arguments or require the missing pieces of pertinent evidence. Such an outcome may be more appealing to the public at large given sensitivities over fairness in testing and admissions, but the shallowness of the court’s reasoning raises the question of whether the complainants were truly granted a fair day at court, or were instead brushed to one side by a court that was only too eager to deploy proportionality as a means to avoid grappling with difficult questions.

Incremental Changes for Proportionality and Religious Accommodation Law

Despite the court’s consistency in dismissing claims for religious accommodations in public exams, there have nevertheless been notable recent developments in proportionality analysis and religious accommodation law that highlight the potential for some changes in the court’s analysis. Namely, in 2018, the court rendered a landmark decision on conscientious objectors, which found that a statute that failed to provide for alternative civilian service was unconstitutional, in reversal of earlier decisions. Meanwhile, the Daegu and Gwangju high courts have ruled that the refusal to provide religious accommodations for school exams and law school admissions interviews, respectively, amounted to unlawful abuses of discretion. ¹²⁹ The issues reviewed and analytical approaches employed in these cases have relevance for religious accommodation cases, including cases on religious accommodation for public exams.

2018 Constitutional Court Decision on Conscientious Objectors

On June 28, 2018, the court issued a landmark ruling pronouncing that Article 5 of the Military Service Act—which does not provide alternative forms of civilian services for conscientious objectors—is inconsistent with the constitution’s guarantee of their freedom of conscience. ¹³⁰ This decision directly overturned the court’s previous rulings on the conscientious objectors issue in 2004, 2011, and 2013, where the court upheld the constitutionality of the Act. ¹³¹ Under the Act, all eligible Korean men are required to serve in the military. ¹³² Anyone eligible who fails to enlist without justifiable grounds would face criminal penalties, including imprisonment. ¹³³ The Act did not allow alternative civilian service for conscientious objectors. Since the 1950s, approximately six hundred conscientious objectors per year were punished for refusing to enlist in the military on the basis of their conscience. ¹³⁴ Almost all of these conscientious objectors in Korea were Jehovah’s Witnesses. ¹³⁵ While this decision does not directly concern the right to religious freedom because the court decided to frame the relevant right as “freedom of conscience” and not “religious freedom,” ¹³⁶ the questions raised in this case are similar to the ones raised in the test accommodation cases, in that the complainants were individuals seeking a type of religious accommodation—that is, exceptions to a uniformly applied rule that goes against their religious beliefs.

The Court’s Proportionality Analysis

The challenged provision was Article 5 of the Act, which lists the different categories of military service allowed under the Act. Article 5 did not provide for any alternative civilian service exceptions.

The court found that the right restricted is the complainants’ right to freedom of conscience. ¹³⁷ The court noted that the majority of the complainants had objected to military service based on their religious beliefs, such that their freedom of religion was also restricted by the relevant statutory provisions. However, the court went on to state that because acting on religious convictions can also fall under freedom of conscience, and conscientious objection can be motivated by ethical or philosophical considerations other than religious beliefs, the court would review the issue as one of freedom of conscience and not freedom of religion.

In its proportionality analysis, the court identified the purpose of Article 5 of the Military Service Act as “to provide fairness in the burden of military service … effective securement of military resources … and efficient distribution of those resources.” ¹³⁸ The court did not provide any external reference to support such articulation and the language does not appear in the Act itself. ¹³⁹ The court concluded quickly—all in one sentence—that such purpose is justifiable and that Article 5 was suitable to achieve such purpose. ¹⁴⁰

Then, in a radical break with the past, the court held that the current Article 5 failed to meet the necessity test, as allowing alternative forms of civilian service would be less restrictive and yet achieve the same purposes intended by the statute. To arrive at such a conclusion, the Court carefully examined the intended purposes of the statute. First, the Court concluded that alternative forms of civilian service would not hinder the achievement of fairness and equity in military service conscription, if the alternative forms of civilian service were designed to be commensurate with the difficulty and length of serving in active duty. ¹⁴¹ Such argument directly overturned the court’s previous rulings that allowing alternative forms of civilian service would impede fairness and equity in military conscription, based on Korea’s social climate that frowns upon exceptions to conscription. ¹⁴²

The court also found that allowing conscientious objectors to serve alternative forms of civilian service would not hinder achieving the purpose of securing military service resources, given that the number of conscientious objectors—six hundred a year—remained immaterial considering the total draft pool and the total number of service members. The court also added that military capabilities increasingly depended on information-dependent, scientific warfare with less emphasis on the number of men. ¹⁴³ In earlier cases, however, the court had held that despite the changing nature of modern-day warfare, the number of service members remained an important factor in Korea’s military capability and for this reason, providing alternative forms of civilian service would not achieve the purpose of securing military resources in an equally effective manner. ¹⁴⁴

Addressing the argument that it could be difficult to distinguish those objectors that are genuine from those who simply wish to avoid serving, the court suggested having a “neutral committee” composed of members from academia, legal practice, religious communities, and civil society review each request for alternative service. The court also suggested that the government prepare a thorough and rigorous review system to sort out the ill-motivated from genuine conscientious objectors. ¹⁴⁵ Here, the court did not raise issues with the additional costs that would necessarily accompany the less restrictive alternative.

Under the section on proportionality stricto sensu, the court emphasized the significant harm conscientious objectors face together with the capability of alternative civilian services to adequately achieve the purpose of the Act. Departing from its previous discussions of the issue, the court emphasized the loss of dignity of conscientious objectors and highlighted that conscientious objectors, in addition to the punishment of imprisonment, are barred from public service jobs for a certain period of time, may lose previously held jobs, and may lose all government-issued patents, licenses and registrations. ¹⁴⁶

Changes and Continued Problems

There were some notable changes in the 2018 conscientious objectors ruling. First, the court placed greater weight on individuals’ fundamental rights than it had previously in the overturned cases. Whereas in earlier decisions, the government’s interest in preserving national security had easily taken precedence over the rights of the affected individuals, in the 2018 case, the court carefully examined the existence and the extent of the harm to the rights of the conscientious objectors. Additionally, the court expressed skepticism of the empirical validity of the government’s claims that allowing alternative civilian service would harm national security, citing the modest number of persons who have historically refused to enlist and the technology-reliant nature of modern warfare. The court also separated its necessity analysis from its review of proportionality stricto sensu. This allowed the court to conduct a deeper analysis, focusing on assessing the effectiveness of the proposed alternative—permitting civilian service—under the necessity prong and then shifting to weighing the respective individual and national security interests under the proportionality stricto sensu prong. All in all, the court’s reasoning in the conscientious objectors case saw considerable improvements with respect to the problems identified in the previous section.

Nevertheless, significant weaknesses remain in the court’s proportionality analysis. For one, the court’s identification of the purpose of the relevant statutory provisions was still conducted in an arbitrary manner, without consultation of either the statutory text or legislative history, and the purpose so articulated was very general. Such a high level and vague purpose— “securing military resources” ¹⁴⁷—would give the court very wide discretion to find that a less restrictive alternative would or would not achieve such purpose.

More importantly, the court continued in its reluctance to expand constitutional protection for religious freedom. Despite acknowledging that the vast majority of the complainants had objected to military enrollment based on their religious beliefs, the court was at pains to characterize this as a freedom of conscience issue and declined to rule on whether the provisions also breached the complainants’ freedom of religion. This leaves open the possibility that the court may apply a different, perhaps stricter, review standard to constitutional complaints that are squarely rooted in religious belief.

2018 High Court Decision on Religious Accommodation for School Exams

In an administrative law case from 2018, the Daegu High Court ruled that a public medical school’s decision not to allow a Seventh-day Adventist student to take make-up examinations for tests held on Saturday was unlawful. ¹⁴⁸ The school’s internal regulations allowed students to take make-up examinations if they were unable to take a test due to “disease or other unavoidable reason,” and the school’s position was that religious beliefs did not constitute such an “unavoidable reason.” ¹⁴⁹

The lower court ruled against the plaintiff, and, on appeal, the High Court ruled that the school’s action constituted an abuse of discretion. ¹⁵⁰ The High Court noted that, notwithstanding the public interest objectives that could be served by the school’s action, such action still violated the principle of proportionality, especially given the intensity of the disadvantage that would be suffered by the plaintiff, who would be unable to graduate from medical school and become a doctor. ¹⁵¹

To reach this conclusion, the High Court reviewed empirical evidence to assess the credibility of the school’s arguments that allowing religious accommodation for tests would impose a prohibitive burden. The court cited examples of medical schools where all exams were held during the week or where Adventist students were allowed to take the test outside of Sabbath hours. It described measures that these schools had taken to ensure fairness and prevent abuse, such as sequestration during holy hours, written commitments not to leak test questions from students who were allowed to test early, and caps on the highest grade that could be earned by those who availed themselves of alternative testing arrangements. The court reviewed the school’s actual practice with respect to test scheduling and administration, noting that whereas the vast majority of examinations (76 percent) had been held on Saturday during the plaintiff’s first semester, the school did not hold any exams on Saturday pending the outcome of the lower court case, and resumed conducting tests on Saturday after the decision came out in favor of the school. ¹⁵² Based on such evidence, the High Court found that there was no prohibitive burden associated with the school’s conducting the tests on a day other than Saturday.

The court’s proportionality analysis was quite abbreviated; rather than sequentially going through the four steps of traditional proportionality analysis, the court simply (1) summarized the government’s arguments on costs and burdens; (2) recognized that those costs and burdens did exist; and then (3) concluded that nevertheless, these costs and burdens were not prohibitive and cost-based considerations were outweighed by the serious harms the plaintiff would suffer. Although the High Court did not employ any groundbreaking theoretical approach to proportionality analysis, its thorough examination of the evidence and close scrutiny of factual assumptions shows how careful consideration of empirical evidence can in and of itself help to improve the quality of reasoning in proportionality cases.

Gwangju High Court Case

In 2021, the Gwangju High Court ruled that a law school unlawfully abused its discretion by refusing a religious accommodation request regarding participation in admissions interviews. ¹⁵³ All Korean law school interviews are held on Saturday, and the plaintiff, who was Seventh-day Adventist, had requested the defendant law school to have her interview scheduled last so that it would take place after sundown. The law school had rejected the request on the basis that the order of interviews should be determined randomly and it would be unfair to make an exception for the plaintiff. The High Court’s decision was appealed by the law school and is pending before the Supreme Court.

In its statement of the legal interests at issue and the standard of review, the High Court opined that although Constitutional Court jurisprudence has distinguished between “internal” and “external” aspects of the freedom of conscience and held only the internal freedom of conscience to be an absolute right, the “non-absolute” status of the external freedom of conscience (that is, the freedom to act on one’s conscience through external actions) does not grant the government unrestricted authority to impose any desired limitations on such freedom. ¹⁵⁴ Courts must therefore strictly scrutinize any restrictions on the external freedom of conscience for compliance with the proportionality test and also assess whether such restrictions violate any essential aspect of the freedom. The High Court further noted that because all law school admissions interviews in Korea were held on Saturday and the situation was unlikely to change in the future, the dilemma the plaintiff faced was a serious one: either violate deeply held religious convictions or give up on the plaintiff’s dream of becoming a lawyer. In essence, the High Court, unlike the Constitutional Court in the Constitutional Court’s test accommodation cases, accorded much weight to the degree of interference with the plaintiff’s constitutional right.

In its proportionality analysis, the High Court acknowledged the law school’s legitimate interest in ensuring the fairness of its admissions interviews process, including by “blinding” the interviews to maintain impartiality, and ruled that the school’s admissions policies satisfied both the justifiability of purpose and suitability tests. However, the school should have afforded “serious consideration” to whether there were alternative measures that would have allowed the plaintiff to participate in the interview without violating her conscience, while preserving the fairness of the process. Conducting a detailed factual analysis of the admissions process, the High Court assessed whether the process implemented by the defendant was the least restrictive alternative. It affirmed the importance of both actual and perceived fairness in the admissions process, and acknowledged that the school should not be required to expend excessive costs, time, or effort in providing an accommodation; thus, any proposed alternative would need to effectively address fairness concerns and not unduly burden the school. The court then found that the school could have scheduled the plaintiff to interview last after sundown while taking necessary steps to maintain fairness, and that the administrative burden and costs associated with this accommodation would have been minimal. Consequently, the school’s refusal to provide the accommodation did not satisfy the necessity test. Furthermore, the measure failed the proportionality stricto sensu test, since the plaintiff suffered grave harm from the school’s refusal to provide an accommodation that could have been granted without raising significant fairness concerns. ¹⁵⁵

Summary

In the above cases, the courts have taken a different approach to the notions of fairness and equality in the context of one’s manifestation of their personal beliefs. Providing for “commensurate accommodation” seems to satisfy the goals of fairness and equality in these decisions, while the incidental costs of providing such accommodation were assessed to be much less important than in the Constitutional Court’s test accommodation decisions.

At the end of the conscientious objectors ruling, the court stated that “to listen to the voices of ‘minorities’ and to reflect their voices is to realize the true spirit of democracy whose core values are tolerance and diversity.” ¹⁵⁶ Similarly, the Gwangju High Court stated in dicta that “[c]onsidering the strength of our country and its global standing, as well as the level of our citizens, at least under the circumstances of this case, sufficient conditions have been established to embrace and accommodate the plaintiff, who is a minority.” ¹⁵⁷

On the other hand, the Constitutional Court has maintained its long-standing position rejecting the right to religious accommodation for public exams. Ultimately, a shift in public opinion in favor of such accommodations may be needed before the court is willing to seriously reconsider the existing jurisprudence. The situation surrounding the court’s conscientious objectors ruling lends support to this view. There is an argument to be made that the court’s reversal of its decades-long position was impacted by national politics and changes in public opinion. While the judiciary is theoretically independent from the administrative and legislative branches, there is a perception among the Korean public that the conscientious objectors ruling was influenced by the tone set by the center-left administration of then president Moon Jae-In, who was previously a human rights lawyer and had spoken in favor of allowing accommodation for conscientious objectors. ¹⁵⁸ A Gallup poll taken in May 2018 (shortly before the Constitutional Court decision was issued) showed that 73.4 percent of respondents were in favor of allowing an alternative means of service for conscientious objectors, even as 66.8 percent said they disagreed with conscientious objection. ¹⁵⁹ This was a significant change from 2016, when another survey found that 53.6 percent of respondents were opposed to an accommodation for conscientious objectors. ¹⁶⁰ The court’s decision also followed resolutions by the United Nations Human Rights Committee denouncing the imprisonment of conscientious objectors. ¹⁶¹ Thus, in a sense, circumstances were ripe for a bold reversal of the court’s earlier position on conscientious objectors.

The test accommodations issue has received far less media attention and such public opinion as does exist appears unfavorable. ¹⁶² This may account in part for the court’s continued reticence to overturn precedent. Nevertheless, the Daegu and Gwangju high court decisions and the dissenting opinion in the 2023 AN Exam Case show that at least some judges are becoming receptive to manifestation of religion claims in general, and requests for religious accommodations in public exams in particular.

Suggestions for Improvement

As the preceding analysis suggests, the Korean Constitutional Court should focus on two main areas to improve its use of proportionality in the test accommodation cases. First, the court should explicitly articulate and safeguard the priority of fundamental rights. In the test accommodation cases, the court was largely silent on the value of fundamental rights and dismissive of the harm to the right at issue, which resulted in a failure to adequately account for the intensity of interference with the right and an analytical approach that was unabashedly utilitarian and majoritarian. While this in part stems from problems inherent in proportionality, the court should still strive to thoroughly assess the value and harm in question instead of continuing with its superficial analysis. If the right concerned is the right to expression of religious beliefs—as the court has articulated in the test accommodation cases—the court should review, at a minimum, the history of religious freedom and expression, including the contours of such rights as intended by the Constitution’s drafters, ¹⁶³ and carefully consider the details involved in the right and harm in question, such as the meaning of Sabbath or holy day.

Second, the court should employ greater methodological rigor in applying the proportionality structure. For one, the court should more clearly distinguish between the necessity and proportionality stricto sensu prongs. The existing jurisprudence heavily emphasizes the necessity prong, which primarily focuses on evaluating suggested alternatives to see if they achieve the desired objective with a comparable level of efficiency as the relevant government measure. Consequently, the proportionality stricto sensu analysis is often treated as a mere formality, leading to a result that is essentially identical to the conclusion drawn from the necessity analysis. However, a measure’s satisfaction of the necessity prong should not automatically imply satisfaction of the proportionality stricto sensu prong. By applying the proportionality stricto sensu test in a more rigorous manner and meticulously assessing whether the public interest at issue outweighs the private right that is infringed, the court could achieve a more comprehensive safeguarding of fundamental rights and strengthen the overall framework of proportionality analysis.

The court should also be conscious of the arbitrariness inherent in defining the purpose of a statute and endeavor to be faithful to the legislative language and history. The court has generally adopted a lenient interpretation of the underlying purpose of relevant government measures, which allows the measure to easily fulfil the first two prongs of the proportionality test. Although this approach may be suitable for the majority of cases, certain cases may involve opaque or multiple purposes and the framing of the issue (i.e., defining the purpose of the statute) may be dispositive.

Relatedly, the court should demand and assess actual evidence to sound out various arguments advanced by the complainants and the government so that the proportionality analysis does not end up as a complete deference to the executive body. Instead of simply accepting the government’s argument that accommodations such as make-up exams or sequestration until the end of holy hours will be administratively inconvenient or raise fairness concerns, the court should engage in a thorough review of the associated risks and their significance. By undertaking a detailed assessment of the nature and magnitude of the relevant costs and risks, the court can ensure a more robust and informed decision-making process.

Conclusion

The Korean Constitutional Court’s cases on religious accommodation for qualifying examinations have revealed important problems in its proportionality analysis—some inherent in proportionality and others arising from or exacerbated by the court’s particular application. Two main areas of focus would improve the court’s use of the proportionality test in religious accommodation cases: (1) an explicit prioritization of fundamental rights, and (2) enhanced methodological rigor in conducting proportionality analysis. In short, the court should restore itself as the guardian of the constitution and fundamental rights. Without these improvements, the court would be relegating itself into a fancy rubber-stamper.

The lack of religious accommodation for qualifying and entrance exams remains an unresolved issue that continues to obstruct members of various religious groups from entering certain professions. According to one study on religious freedom issues impacting Seventh-day Adventists, between 2015 and 2020, more than six hundred Seventh-day Adventists were negatively affected by the lack of religious accommodations in forty-six different types of public exams. ¹⁶⁴ Public agencies have cited the court’s test accommodation jurisprudence when refusing requests for a religious accommodation. ¹⁶⁵ Members of other Christian denominations have also faced such challenges in connection with exams being scheduled for Sunday. Having failed to procure judicial remedies, some have engaged in activism to solve the problem through lobbying or legislation, at times to the detriment of other religious groups. ¹⁶⁶ Improving the court’s proportionality analysis would give those who seek religious accommodation a fair day at the court and potentially, a constitutionally protected chance at realizing their life-long dreams.