Courts in the Global South and rights protection

Courts’ active role towards rights protection is a pattern that has surfaced in Latin America and other countries of the Global South. Judges intervene in key structural cases of broad violations of socioeconomic rights, keeping public and private parties accountable and enforcing structural remedies (Rodríguez-Garavito Reference Rodríguez-Garavito2010). In one example of this pattern, the Supreme Court of India has addressed environmental issues (Sahu Reference Sahu2008) and nonjusticiable social rights, and in some cases has elevated them to the status of fundamental rights (Shankar and Mehta Reference Shankar, Mehta, Gauri and Daniel2008). The Constitutional Court in South Africa has had a key role in cases of housing and health care, engaging the government in actions to advance rights protection (Pieterse Reference Pieterse2008). In Latin America, the Costa Rican Supreme Court Constitutional Chamber has relaxed procedural rules and encouraged citizens to file suits, empowering marginalized groups to protect their rights (Wilson and Cordero Reference Wilson and Cordero2006). In Brazil, courts have shaped how the right to health should be protected (Ferraz Reference Ferraz2010). The Colombian Constitutional Court has led one of the most far-reaching rights revolutions in Latin America (Wilson Reference Wilson2009).

Whether the impact of these courts’ rulings is far-reaching has been assessed, with mixed results. In South Africa legal developments to protect socioeconomic rights have not brought social transformation (Langford Reference Langford, Langford, Cousins, Dugard and Madlingozi2014). In Chile the Supreme Court did not promote rights protection and chose a role that would allow it to protect its status among other political actors (Couso Reference Couso, Sieder, Schjolden and Angell2005). In contrast, Costa Rica and Colombia have led a “real rights revolution” where broad definitions of legal standing and low cost in the access to higher courts allowed individuals to successfully use litigation to protect their rights (Wilson Reference Wilson2009). Ecuador, Honduras, Paraguay, and Venezuela offer examples of courts that are perceived as captured by political influence and having minimal effects on rights protection (Sanchez Urribarri Reference Sanchez Urribarri, Dixon and Ginsburg2017).

Researchers have explored whether the type of right under litigation, among other factors that affect courts’ decision-making, leads to different outcomes. Simmons (Reference Simmons2009) found that in litigation under international treaties, when these treaties focus on a right that affects the government’s ability to achieve its political goals the treaty will be less effective. In the case of the Brazilian Supremo Tribunal Federal, justices largely support mainstream government policies and are less likely to support second- and third-generation rights, although they have been cautiously moving toward a more active approach (Brinks Reference Brinks, Helmke and Ríos-Figueroa2011). In Costa Rica and Colombia, LGBT communities have achieved positive outcomes when litigating in anti-discrimination claims, but when focusing on same-sex marriage or adoption by same-sex couples, they have not been successful (Wilson and Gianella-Malca Reference Wilson and Gianella-Malca2019). In cases of abstract constitutional review, the Colombian Constitutional Court defers significantly more often to the government in cases involving the protection of individual rights than when individual rights are not at stake (Rodríguez-Raga Reference Rodríguez-Raga, Helmke and Ríos-Figueroa2011). In freedom-of-speech litigation in Colombia, India, and South Africa, rulings vary depending on the content of the claim: in sexually explicit speech, courts tend to rule in favor of parties with resources; in cases of economic speech, judges rule in favor of nonelite claimants; while in cases of religious speech, results are mixed (Botero et al. Reference Botero, Ellett, Keck and Stohler2021). Considering this literature, I expect the type of right under litigation to affect who wins in AP.

Party capability theory

Another factor that shapes judicial decision-making and has been broadly explored in the literature is party capability. Parties with more financial resources and litigation experience (“the haves”) are more successful in litigation than those with fewer resources and less experience (“the have-nots”) (Galanter Reference Galanter1974). Researchers have emphasized the advantages the government holds when acting as a plaintiff or a defendant in litigation (Wheeler et al. Reference Wheeler, Cartwright, Kagan and Friedman1987). Kritzer (Reference Kritzer, Herbert and Susan2003) found that in the United States the government has a fundamental advantage in litigation because rules are made by the government and judicial workers are, in the end, governmental workers.

The applicability of party capability has been explored in different legal systems and countries, with mixed results. In England, Atkins (Reference Atkins1991) found evidence that party capability is one of the factors explaining the decision-making process of the Court of Appeals and that asymmetries between individuals and the government are strong predictors of the court’s decisions. Xin He and Yang Su (Reference He and Su2013) analyzed adjudication decisions from Shanghai courts and found that stronger parties are more likely to win in litigation, but given the strong influence of the government over the judiciary, it is impossible to ensure equal footing for parties in litigation.

Studies in Israel and Taiwan have found limited evidence of the applicability of party capability. In Israel, the “haves” enjoy limited advantages in litigation outcomes; other factors like legal representation, judges’ ideology, and the ability to reach favorable out-of-court settlements bring advantages to “weaker parties” (Dotan Reference Dotan1999; Dotan and Hofnung Reference Dotan and Hofnung2001). In Taiwan, there is evidence that better legal representation and the court’s discretion over its docket have larger effects than party resources (Chen, Huang, and Lin Reference Chen, Huang and Lin2014).

In the Colombian case, evidence conflicts on how party capability affects judicial decision-making. Páez-Murcia, Lamprea-Montealegre, and Vallejo-Piedrahita (Reference Páez-Murcia, Lamprea-Montealegre and Vallejo-Piedrahita2017) show that in AP cases of administrative morality the government has won more cases than other parties, but in AP cases of environmental litigation individuals (who are litigants with fewer resources) have obtained favorable rulings more frequently than the government. Botero et al. (Reference Botero, Ellett, Keck and Stohler2021) found that in cases of freedom of expression, the Constitutional Court rules differently in favor of elites or in favor of nonelite actors, depending on the content of the case. Considering the literature, I expect for the government to achieve positive outcomes in AP litigation in some areas, but I also expect individual plaintiffs to come out ahead in AP in other areas.

The Colombian context: The 1991 Constitution, collective litigation, and the council

The 1991 Constitution included a rich enunciation of constitutional rights (Wilson Reference Wilson2009), and one of the new writs to protect them is AP. AP is a cause of action to enforce collective rights (Art. 88), and any person or organization has legal standing to file one (Congreso de la República de Colombia, Ley 472, 1998). The constitution and the law provide a long list of collective rights (Constitución de Colombia 1991, Art. 88; Ley 472, 1998; Ley 1437, 2011) that can be categorized in two groups: first, collective public goods that do not ensure individual compensation (e.g., public patrimony, public space, public security, and administrative morality) (Art. 86, Constitución de Colombia; Art. 4, Law 472 of 1998); second, goods that although collective in nature have led to individual compensation (Art. 86, CP; Art. 4, Law 472 of 1998). Examples of this category are public health and environmental cases where the council has required the government and other private parties to provide specific goods and services to citizens (Rodríguez-Garavito Reference Rodríguez-Garavito2010).

Collective litigation at its core fulfills the need of leveling the ground in litigation between the “haves” and the “have nots.” During the constitutional debates, the framers addressed existing inequalities in litigation and designed AP to face this problem ^{Footnote 2} under the assumption that individual citizens would mobilize to protect collective rights: “It [AP] has been an important advance towards making a more solidary society … so the only way to protect rights [collective constitutional rights] is through the actions of one or more individuals to protect the group” (Asamblea Constituyente Comisión Quinta, April 16, 1991, 9). Some framers expressed concern that this writ could raise the standards of governmental responsibility, making it liable for risks and damages beyond its control. ^{Footnote 3} Despite these tensions around the framing of AP there were high hopes for the social changes that it could bring.

The Council of State is the administrative court of highest hierarchy in the country and the one with jurisdiction over AP. The role of the council in rights protection is part of the mixed system of judicial review designed by the 1991 Constitution. The constitution followed a concentrated approach by granting the Constitutional Court jurisdiction to review laws and executive orders; also, it granted the council jurisdiction to review administrative regulations that do not fall under the jurisdiction of the Constitutional Court (Arts. 237 and 241, Colombian Constitution). The constitution followed a diffuse system in two ways: (1) it allows all courts in the country to declare the inapplicability of statutes deemed contrary to the constitution; and (2) the constitution created writs like tutela and AP (Brewer-Carías Reference Brewer-Carías2009) that allow for immediate protection of constitutional rights to be granted by judges. Other countries in Latin America that adopted a mixed system of judicial review are Brazil, Dominican Republic, Ecuador, Guatemala, Mexico, Nicaragua, Peru, and Venezuela (Navia and Ríos-Figueroa Reference Navia and Ríos-Figueroa2005; Brewer-Carías Reference Brewer-Carías2009)

Justices at the council have recognized the challenge they face to meet the expectations of the 1991 constitutional reform. During debates on the Administrative Code in 2009, justices at the council argued specifically that administrative judges could also be “rights protectors.” Traditionally administrative judges have been perceived merely as enforcers of laws, but during these debates justices expressed the need for them to embrace a new role, closer to the one that the Constitutional Court has performed (Exposición de Motivos, Ley 1437, 2009).

In this article I focus on four collective rights: environmental protection, public security, the rights of consumers, and administrative morality. I chose these rights because they target different policy areas and aspects of social life; I am aiming to capture variation given their different goals and remedies. There are also noticeable differences in the level of use of these four rights: environmental protection has been the most frequently used between 2010 and 2018 in the country (15%); consumers and users and administrative morality have been used less frequently (5% and 8% respectively), while public security has been used in less than 1% of the suits filed (Londoño-Toro and Torres-Villarreal Reference Londoño-Toro and Lucía Torres-Villarreal2012). These data are taken from a sample of 3,639 AP cases in the country, which represents only 3.2% of the total AP cases filed. Although these are not enough data to draw conclusions about patterns in Colombia, these are the cases that the Ombudsman Office keeps in the only centralized database of AP cases in the country, which highlights the limited availability of data on AP.

AP has acted as a fire alarm mechanism when the government or private parties are harming or misusing resources that should benefit society at large or specific communities. Although collective rights are not redistributive in nature (e.g., they don’t assign resources directly to lower-income populations), they aim to improve situations where public goods are at stake. These situations will benefit the public in general and in particular those who are more in need of public resources because they do not have access to privatized versions of these services (e.g., public space and public security). From this perspective AP protects equitable access to governmental services.

Expectations and hypotheses

I expect to find that the Council of State’s decisions per type of right are significantly different. I will test this hypothesis with a t-test (Appendix 3):

I also expect to find that both party resources and type of right have an effect on the council’s decision-making, even taking the other into account along with year as a control and an indirect measure of other variables. I will test these expectations with the following two hypotheses using a logistic regression (Appendix 3):

$$Decision = {B_6}\left( {{P_{res}}} \right) + {B_5}\left( {{D_{res}}} \right) + {B_4}\left( {{R_{Env}}} \right) + {B_3}\left( {{R_{Sec}}} \right) + {B_2}\left( {{R_{Cons}}} \right) + {B_1}\left( {Year} \right) + {B_0}$$

Data and methods

In exploring whether party resources and the type of right affect the councils’ decision-making I analyze an original database of all the cases of AP decided by the council at the appeal level from 1997 to 2017. I analyzed 516 cases in total. ^{Footnote 4} I also analyze interview data with relevant actors in AP.

I focus on cases at the appeal level for two reasons: first, there is no database that centralizes information of the courts in all thirty-two departments, so no reliable information at the department level is available. ^{Footnote 5} Second, council decision-making affects other courts in the country that use the council’s jurisprudence to support their decisions. ^{Footnote 6} Other studies in party capability, both in the United States and in other countries, have similarly focused on courts at the national level (Songer, Sheehan, and Haire Reference Songer, Sheehan and Brodie Haire1999; McGuire Reference McGuire1995; Dotan Reference Dotan1999; Chen, Huang, and Lin Reference Chen, Huang and Lin2014).

I identified the cases as they are available on the council’s web search engine by using a variety of key terms (Appendix 1). ^{Footnote 7} I used additional criteria to clean up my data: I did not include cases that were filed through other injunctions (e.g., torts, public contracting, and judicial review). I focused on final rulings (sentencias) and did not include procedural decisions (autos); I assumed that only in the final ruling the council decides on who wins the case. I only included cases that could fall under one type of collective right. Although plaintiffs can base their claims on the violation of one or several constitutional rights, I decided to exclude cases with overlaps to be able to explore the effects of individual rights (Appendix 2 lists these cases). I focused on cases where the keyword matched the core legal claim. After applying these criteria, I identified 510 relevant cases.

My key variables are parties’ resources and the core legal claim. I coded information on the types of parties and assigned them to broader groups. Files of the cases do not include specific information related to the parties’ socioeconomic status, and only some cases mention whether parties had legal representation. Instead, I followed what has been done in other studies (Wheeler et al. Reference Wheeler, Cartwright, Kagan and Friedman1987; Songer, Sheehan, and Haire Reference Songer, Sheehan and Brodie Haire1999; Songer and Sheehan Reference Songer and Sheehan1992) by assigning plaintiffs and defendants to broader groups and then making assumptions of their resources. Although this “party typology measure” has been criticized (Szmer, Songer, and Bowie Reference Szmer, Songer and Bowie2016), given the limitations in available data I conducted a qualitative analysis of these cases and collected interview data with relevant actors that allowed me to gather insights into party resources.

I used six groups to categorize plaintiffs and defendants: individuals; nonprofits; private businesses; local governments; department governments; and the national government. Under the nonprofit category I used different codes for local and transnational nonprofits; no transnational nonprofits appeared in the data. I used different codes for big businesses and small businesses, and coded this information using web searches of businesses’ budgets and range of their operations (Appendix 4). Regarding individual litigants, I considered whether the individual was acting on her behalf or representing an organization. I coded these cases accordingly. Following conventions in the literature I assumed that (1) individuals have fewer resources than nonprofits, businesses, or the government; and (2) nonprofits and small businesses have fewer resources than large businesses.

I also coded information on the type of right under litigation as it was identified in the suit. The variable “type of right” was coded following the criteria I described earlier in this section (see also Appendix 2).

I included a control variable for the year when the decision was made. The literature on constitutional reforms in Latin America suggests that factors like domestic political dynamics (Sanchez Urribarri Reference Sanchez Urribarri, Dixon and Ginsburg2017; Bernal Reference Bernal, Dixon and Ginsburg2017), ideology, and the structure of the judiciary (Landau Reference Landau, Dixon and Ginsburg2017; Vieira Reference Vieira, Dixon and Ginsburg2017) have shaped rights protection. Tracking time is a way to explore how the council has reacted to these factors; this variable will show whether there are shifts in the council’s decision-making over time that could correspond to key political or ideological changes in Colombia. My dataset includes cases from 1997 (when the first case was decided) through 2017.

Regarding the council’s decision, I followed what has been done in other studies, coding who won the case by comparing what was requested in the suit and whether the court granted it or not (Songer, Sheehan, and Haire Reference Songer, Sheehan and Brodie Haire1999; Haynie Reference Haynie1994). ^{Footnote 8} The variable “decision” was coded as a win for the plaintiff if the council declared that the constitutional collective right was violated and granted the injunction. If the council stated that no violation of collective rights took place, then it was coded as a win for the defendant. In a few cases (n = 6) the council partially granted the case to the plaintiff, for instance by stating that a constitutional collective right was violated but that the defendant was not responsible; I coded these cases as a tie and did not include them in the analysis.

Cases on administrative morality were coded entirely by the researcher. Research assistants provided support in coding cases on public security, environment, and consumers’ protection.Footnote ⁹ In these cases, a sample of cases was coded independently to assess intercoder reliability. In each case, a minimum of 91% reliability was achieved, showing that the coding was highly reliable. ^{Footnote 10}

I conducted thirteen interviews with key actors in Colombia, including plaintiffs, bureaucrats in public organizations, justices at the council, and citizens who have followed the implementation of these rulings. ^{Footnote 11} Interviewees were selected on the basis of their experience with AP as litigants or their experience with the implementation of rulings. I identified main narratives in these interviews and coded them accordingly.

Findings

Plaintiffs in these cases are individuals in a vast majority (82.6%, n = 426; most of them men), while most defendants are local governments or local agencies (62.7%, n = 322). In a few cases, governmental agencies have appeared as plaintiffs (5.24%, n = 27), particularly when oversight agencies, like the Ombudsman, have filed suits (Figure 1).

Who wins in these cases varies considerably depending on the type of right under litigation. While in cases of public security and environmental issues plaintiffs are more likely to win, in cases of administrative morality defendants are more likely to win (Table 1). Even when comparing cases of environment and public security, the two categories in which the plaintiffs’ likelihood to win is high, the difference is significant (Table 2).Footnote ¹²

Table 1. Likelihood of plaintiffs winning cases, by type of right.

Table 2. Two sampled t-test, comparisons of likelihood of litigants winning cases by right. Plaintiff win 0/1.

Litigation in administrative morality is a scenario where plaintiffs tend to lose in the highest proportion in comparison to the other rights. The council has stated that a violation of administrative morality is the illegal behavior of a public agent, in which the motivation was different from the public good (AP170 of 2001; 68001-23-31-000-2011-00148-01). Litigation on administrative morality represents 30% of the cases in the database, which shows that although this right is frequently used, the council is not supportive of these claims.

Conversely, in cases of public security and environmental protection citizens have prevailed in litigation over the government. The council has acknowledged that the government is responsible for providing safe conditions that will allow citizens to enjoy life in community in public spaces such as roads, parks, and sidewalks (e.g., case 68001-23-15-000-2003-00765-01). In cases of environmental violations, the council has ruled in favor of plaintiffs on the assumption that the government is responsible for protecting the environment and enforcing environmental regulations. An example of this is a case in which citizens sued the City of Paipa for granting a license to a colliery that polluted the river (case 15001-23-31-000 2001-01470-01).

In consumers’ rights, plaintiffs win in 41% of the cases. These cases focus on damages suffered by groups of consumers due to the conditions that have been set by retailers, businesses, and service providers (e.g., case 25000-23-15-000-2010-02799-01). Plaintiffs are required to prove that the damage is collective rather than individual; lack of evidence on this aspect is what motivates rulings against plaintiffs.

Following studies of party capability, I calculated the success rate per type of party (Table 3). ^{Footnote 13} This measure describes whether parties’ resources increase their likelihood to come ahead in litigation. The overall success rate is calculated as the overall average of success when acting as either a plaintiff or as a defendant, but not both. I coded information on businesses’ (small or large) and nonprofits’ (local versus transnational) size and found that no transnational nonprofits appeared in these cases. ^{Footnote 14} Businesses had but a few appearances in both roles, as plaintiffs (one small and two large businesses) and defendants (two small and twenty large businesses). I did not separate these categories in Table 3 because the frequencies were low and the separation along resources became no longer meaningful.

Table 3. Party success rates overall.

The overall success rate shows that the national government wins the most often (63.7% overall success rate), which confirms my expectations that the government holds an advantage in AP. Individuals came in second with a 55.4% win; this pattern is surprising because despite individuals’ limited resources and experience, they come out ahead when facing other parties that are presumably more resourceful and experienced, such as local governments and businesses. It is noticeable that while local governments are frequent plaintiffs and defendants in these cases (they appear 313 times in these cases), their success rates are low when compared to parties with fewer resources.

To further explore these patterns, I calculated parties’ success rates per type of right and found that who wins varies dramatically with the right in litigation (Table 4). For example, in cases of public security and environmental protection, individuals’ success rate is the highest among all party categories. In cases of administrative morality, the government (national, department, and local levels) has the highest success rate among all categories. Although individuals have the highest number of appearances in litigation of administrative morality (125), their success rate is only 21.6%. Additionally, parties engage in litigation differently by type of right. For example, local governments have engaged in environmental litigation more frequently than in other rights.

Table 4. Party success rates by type of right.

In summary, the data in Tables 3 and 4 show that parties with more resources do not always come ahead. The analysis per right shows considerable variation on who wins these cases. To know whether these patterns are statistically significant I modeled the conditions on which parties win, based on their presumed resources and the type of right.

Modeling the council’s decision-making and party capability

I used a multivariate logistic regression model where the dependent variable is “plaintiffs’ wins,” coded as 1 if the plaintiff won the case and 0 otherwise. The independent variables are plaintiffs’ resources, defendants’ resources, type of right under litigation, and year of the decision. The variables plaintiffs’ resources and defendants’ resources are ordinal measures coded on the basis of the party’s presumed budget, resources, and litigation experience. These variables are coded as follows: 1 = individual, 2 = small businesses and nonprofits, ^{Footnote 15} 3 = large business, 4 = local government, 5 = department government, and 6 = national government. I expect plaintiff’s resources to be positively correlated to the dependent variable and defendants’ resources to be negatively correlated with the dependent variable. ^{Footnote 16}

I created dummy variables to control the effect of the type of right (environment, public security, or administrative morality). Finally, the variable “year” is a scale that ranges from 0 to 20 (decisions range from 1997 to 2017). In Appendix 3 I describe the hypotheses tested by the model and their interpretation.

The model confirms that plaintiffs’ and defendants’ resources are significant (at the .05 and the .01 levels respectively) and in the expected direction. Plaintiffs’ resources are positively correlated to plaintiffs’ wins, while defendants’ resources are negatively correlated to plaintiffs’ wins. The coefficients for each variable are shown in Table 5.

Table 5. Logistic regression. Analysis of likelihood of plaintiff success in AP in environmental protection, administrative morality, public security, and consumers and users at the council.

N = 506.

Pseudo R² = .1999.

Control: Right: Administrative Morality (omitted from the table).

Dependent variable: plaintiff win 0/1.

⁺ Independent variable explained in text.

*p < .05; **p < .01; ***p < .001.

The dummy variables “right: environment” and “right: security” are significant at the .001 level and both are correlated with an increased probability of a favorable ruling for the plaintiff compared to administrative morality. This is evidence that plaintiffs are more likely to win when filing suits on public security and environmental issues than in administrative morality. The variable “right: consumers” is not significant when compared to administrative morality so there is no evidence to suggest plaintiffs are more or less likely to win, after controlling for the other variables.

I also explored whether the effects of parties’ resources vary depending on the type of right under litigation. I ran a second model where I tested for interactions and found that the variables of the interactions were not significant. Although my data does not provide evidence that parties’ resources mediate the effect of the type of right, there are indications that type of right mediates party resources, as shown with the t-test in Table 2. A possible explanation for these results is that when breaking down the data into clusters of type of party versus type of right (e.g., cases where nonprofits have appear in litigation in cases of administrative morality) the number of observations in my database is small. More data will be needed to further explore this relationship.

The model shows that “year” is a strong predictor of the council’s decisions (p < .001) and is positively correlated to plaintiffs’ wins, which indicates that over time the council has tended to rule in favor of the plaintiffs. I found that the growth in plaintiffs’ wins over time has been gradual, and the changes in the data do not coincide with elections or other major political events in the country (Figure 2). ^{Footnote 17}

Figure 1. Plaintiffs and defendants in AP cases for environmental protection, administrative morality, public security, and consumers’ rights (author’s own data). Plaintiffs n = 516. Defendants n = 514.

Figure 2. Overall plaintiff success rate by year. N = 507.

Analysis

My findings show that both type of right and party capability affect how the council rules. From the perspective of the type of right, plaintiffs are more likely to win in environmental and public safety cases. From the perspective of type of party, the national government is more likely to reach a positive outcome than any other party, and individuals have higher rates of success than stronger parties such as department and local governments. In the following pages I will unpack these findings.