X Factors and Tipping Points in Eviction Cases: A Statistical Analysis of Eviction Litigation of the European Court of Human Rights

Abstract

This study is the first to statistically analyse all eviction case law of the European Court of Human Rights up to 2024. It examines the impact of various case characteristics, such as the type of tenure, the reason for eviction and the involvement of vulnerable people, with a specific focus on the role of ownership and property rights. The study evaluates whether evicted homeowners have a higher success rate compared to non-owners, as they can invoke both Article 8 and Article 1 of Protocol No. 1. Additionally, it investigates whether landmark cases like McCann v the United Kingdom and F.J.M. v the United Kingdom serve as ‘tipping points’ that significantly altered the Court’s jurisprudence. The findings highlight the substantial influence of the McCann case on subsequent eviction rulings and reveal that being a tenant in the public rental sector significantly increases the likelihood of the Court finding a violation of Article 8.

1. INTRODUCTION

National and international (human rights) law are essential to protecting people against eviction—the forced removal from one’s home—and homelessness. While numerous countries have ratified the same human rights treaties, research shows that the level of protection varies significantly across jurisdictions.¹ In an attempt to set a minimum level of protection against eviction in the (then) 47 Member States of the Council of Europe, the European Court of Human Rights (the Court) ruled in 2008 that ‘any person’ at risk of losing their home should be able to ‘have the proportionality of the measure determined by an independent tribunal in the light of the relevant principles under Article 8’ of the European Convention on Human Rights (the Convention), which protects the right to respect for the home.² This minimum level of protection has mainly procedural value as it does not address the proportionality assessment itself or the threshold at which an eviction violates the right to respect for the home. Nevertheless, the case in which the Court made this ruling, McCann v the United Kingdom, is considered a landmark case which brought the proportionality test to the forefront in both case law and scholarly literature.³

For a long time, the Court held on to the McCann-minimum level of protection.⁴ As such, both national courts and legal scholars were under the impression that a proportionality assessment should be seen as a mandatory step in any eviction procedure if the evictee would raise such a defence.⁵ Vols even argues that McCann caused a ‘push towards a context-sensitive adjudication in eviction litigation’.⁶ Despite this impact, the Court mitigated the effect of McCann about a decade later. In 2016, the Court ruled in Vrzić v Croatia that the approach towards the proportionality assessment could be different in cases where two private interests are at stake (e.g. between an evictee and a private landlord or mortgage lender) as opposed to cases where only the private interests of the evictee are at stake (e.g. in case of State-owned or socially owned housing).⁷ The Court confirmed this reading in 2018 in the case F.J.M. v the United Kingdom and ruled that a proportionality assessment is not automatically required in cases where a private individual seeks eviction. It emphasises that the applicants in both cases (Vrzić and F.J.M.) voluntarily entered into contractual relationships whereby they risked losing their homes, which diminishes the role of the Convention.⁸

Legal scholars have frequently discussed whether these rulings depart from McCann and whether they imply that an eviction initiated by a private property owner is exempt from proportionality testing.⁹ While most agree that Vrzić and F.J.M. will impact the legal protection offered against evictions initiated by private entities, they also agree that these rulings do not necessarily mean that proportionality defences are not allowed in such cases by default.¹⁰ In fact, some even argue that the Court misinterpreted its previous decisions in the F.J.M. case. They claim that property rights and ownership should just be factors to be considered in a proportionality analysis.¹¹ Nonetheless, both rulings clearly show that the Court places considerable emphasis on property owners’ rights and interests in eviction proceedings.

It is no secret that the right to property stands out among the strongest rights, especially compared to personal rights such as the right to respect for the home.¹² The mere existence of eviction as a legal instrument demonstrates this power dynamic.¹³ However, this does not mean that ownership always trumps non-ownership in eviction cases. Evictions are not automatically justified when two private parties have a contractual relationship. Moreover, the presumptive strength of ownership does not mean that owner-occupiers are mechanically successful in preventing their eviction by invoking both the right to housing and the right to property. This raises questions about the actual impact of ownership in eviction cases and whether this influence has changed since the pro-property law rulings in the Vrzić and F.J.M. cases.

In this article, I will conduct a statistical analysis of all eviction cases adjudicated by the European Court of Human Rights until 2024, totalling 200 cases. The objective of this article is to unveil the influence of ownership on the outcome of these cases and to explore the impact of various other contextual factors, such as the type of tenure, the reason for eviction and the involvement of vulnerable people. If I would write this in a formula, my aim is to discover the X factor(s) in the equation: wrongfully evicted = eviction + X. While eviction itself is already recognized as a severe intrusion upon the right to respect for the home, it is this enigmatic ‘X factor’ that elevates an interference into a full-fledged violation and makes the Court decide that the applicant has been wrongfully evicted.

What is more, this article assesses whether landmark cases such as McCann, Vrzić and F.J.M. can be identified as ‘tipping points’ that brought about significant changes in the Court’s jurisprudence. Tipping points refer to ‘moments in time where a normally stable or only gradually changing phenomena suddenly takes a radical turn’.¹⁴ I examine whether these cases or other rulings have had such a significant impact on case outcomes. Identifying and analysing tipping points helps to explain the extent to which individual cases influence the outcomes of future cases.

There is a growing amount of scholarly literature that examines the issue of eviction and the protection provided by the Convention. However, the majority of this literature tends to focus on specific aspects, such as the interpretation of Article 8 in a particular country,¹⁵ the relevance of a specific case,¹⁶ the application of the proportionality principle,¹⁷ or they focus on either evictions initiated by private entities (private evictions) or evictions instigated by state authorities (public evictions).¹⁸ These studies have revealed the highly contextual nature of the Court’s jurisprudence.¹⁹ However, they also reveal a discernible gap in our understanding regarding the influence of individual circumstances on the Court’s decisions. By conducting the first statistical analysis of all eviction cases adjudicated by the Court, this article aims to fill this gap by providing novel quantitative insights into the relationship between such circumstances and the case outcome.

This article is divided into four sections, including this introductory section. The second section describes the level of protection that is offered against eviction by discussing the case law from the Court. This section focuses on the role of the right to housing and the right to property in eviction cases. Section 3 discusses the data and quantitative methods used and presents the results of the quantitative case law analysis, with a focus on potential tipping points, X factors and the role of the right to property. The final section includes the discussion and conclusion.

2. PROTECTION AGAINST EVICTION: THE RIGHT TO PROPERTY AND THE RIGHT TO HOUSING

In most eviction cases, at least two human rights play an important role: the right to housing and the right to property. If a tenant is evicted by a private landlord, the tenant’s housing right collides with the landlord’s right to property. However, if a homeowner is forced to leave the home, both rights reside with the same person and contribute to the protection against eviction. As such, depending on the situation, the right to property could be seen as an enemy to the right to housing or an ally.²⁰ This section discusses the relationship between these two rights in the context of evictions and describes the level of protection against eviction that is required by the European Court of Human Rights. Although the Convention does not contain an explicit right to housing, Article 8 of the Convention guarantees the right to respect for one’s home. This provision has been interpreted to include certain housing rights, particularly those offering protection against evictions, which is a fundamental aspect of the right to housing.²¹ The right to property, on the other hand, is explicitly recognised in Article 1 of Protocol No. 1 of the Convention.

According to the rights paradigm as described by Van der Walt, property rights have the force and status of being superior; they stand out as among the strongest rights, especially compared to personal rights, such as the right to housing. In other words, ownership trumps non-ownership.²² Hence, in an eviction case, a property owner would win against those without a legally recognized property right, at least theoretically. In fact, Van der Walt refers to eviction as ‘the landowner’s strongest and most valuable weapon and arguably the most eloquent symbol of the rights paradigm’.²³ However, theory and practice do not always align. Van der Walt acknowledges that this ‘rights paradigm’ does not exist in a pure form and could be challenged or undermined by legislative interventions and judicial interpretation. An example of the former is legislation that protects the interests of tenants against unfair, arbitrary or unlawful evictions.²⁴ An example of the latter is the proportionality assessment of the European Court of Human Rights, in which the personal circumstances and competing interests of the tenant are taken into account.²⁵ Hence, in practice, the outcomes of eviction disputes are usually not (solely) determined by the rights paradigm.

The European Court’s focus on proportionality is often attributed to the landmark case McCann v the United Kingdom. Yet, this judgment does not portray an abrupt shift, but was rather the result of a gradual evolution towards a stronger focus on procedural protection and personal circumstances. In earlier cases like Buckley v the United Kingdom and Chapman v the United Kingdom, dissenting judges began to emphasize the importance of accounting for personal hardship, vulnerability and the limited margin of appreciation afforded to Member States in eviction cases.²⁶ The majority of the Court first confirmed to this reading in Connors v the United Kingdom.²⁷ In Connors, the Court ruled that the statutory framework allowing the applicant and his family to be evicted without a proper hearing failed to meet the standards of necessity and fairness. The absence of judicial oversight and the inability to challenge the eviction on substantive grounds underscored the need for greater procedural protection.

The pivotal ruling in McCann v the United Kingdom marked the culmination of this growing trend. The Court started by ruling that ‘the loss of one’s home is a most extreme form of interference with the right to respect for the home’ as enshrined in Article 8 of the Convention.²⁸ Such interference is only permissible if it is in accordance with the law, if it is necessary to pursue the legitimate aims listed in Article 8, paragraph 2, and if the interference is ‘necessary in a democratic society’. An eviction will be considered necessary in a democratic society if it answers a ‘pressing social need’, if the reasons for eviction are ‘relevant and sufficient’ and if the eviction is proportionate to the legitimate aim pursued.²⁹

In McCann, the Court ruled that any person at risk of eviction should be able ‘to have the proportionality of the measure determined by an independent tribunal in the light of the relevant principles under Article 8 of the Convention, notwithstanding that, under domestic law, his right of occupation has come to an end’.³⁰ This means that if the evictee argues that the eviction is disproportionate, domestic courts must assess whether the eviction is necessary considering the evictee’s personal circumstances and the broader socio-economic context. This assessment typically involves weighing the individual’s right to housing and their personal circumstances against the reason for eviction and the interests and rights of the evictor (i.e. the State or property owner).³¹ Although domestic courts are responsible for making the initial assessment, the European Court has the duty to determine whether they failed to uphold this responsibility.³² Hence, the final proportionality assessment is for the European Court, and it is up to the national authorities to prove the need to restrict the right to respect for the home in a particular case. The Court might, for instance, conclude that the applicant’s right to respect for their home outweighs the state’s legitimate interest in controlling its property.³³

Since McCann, the Court has repeatedly emphasized the importance of procedural safeguards,³⁴ leading to multiple breaches of Article 8 because domestic courts failed to perform a proportionality analysis.³⁵ According to Vols, McCann caused a ‘push towards a context-sensitive adjudication in eviction litigation’.³⁶ Yet, while the proportionality test received ample attention in both literature and case law since McCann in 2008, the contextual case-by-case adjudication is not present or required in every case. First, a proportionality review is only required if the evictee raises an ‘Article 8 defence’, which means that the evictee should raise relevant arguments concerning the (dis)proportionality of the eviction.³⁷ Second, States have a margin of appreciation regarding the assessment of ‘necessary in a democratic society’. The rule of thumb is that the scope of the margin of appreciation is relatively narrow if ‘intimate or key rights’ are at stake, whereas the margin of appreciation is relatively wide if social and economic policies are at issue or if ‘there is no reason to doubt the procedure followed’.³⁸ This means that the scope of the margin of appreciation largely depends on the circumstances of a case, especially in eviction cases.³⁹ This is because the Court repeatedly confirms that the right to respect for the home is crucial to ‘the individual’s effective enjoyment of intimate or key rights’,⁴⁰ but is also often affected by general social and economic policy considerations.⁴¹ Nevertheless, the Court has held in multiple cases that the margin of appreciation in housing matters is narrower when it comes to the rights guaranteed by Article 8 compared to those in Article 1 of Protocol No. 1 to the Convention. This is because the right to respect for the home is central to ‘the individual’s identity, self-determination, physical and moral integrity, maintenance of relationships with others and a settled and secure place in the community’.⁴²

Third, the obligation to conduct a proportionality test might vary based on the type of ‘evictor’. In Vrzić v Croatia (2016), the Court deviated from its previous ruling in McCann by ruling that the approach towards proportionality is different in cases between two private parties as opposed to disputes between the State and a private party, especially if the purpose of the eviction is to protect the rights of others, in this case, a private landlord.⁴³ The applicants in the Vrzić case voluntarily used their home as collateral for their loan and agreed that if they failed to pay their debts, the creditors were entitled to seek enforcement of the repayment through the sale of their house, which is exactly what happened. In other words, the applicants agreed to be evicted if they did not meet their contractual obligations. Therefore, the Court ruled that there had been no violation of Article 8 despite the absence of a proportionality assessment by an independent court.⁴⁴

In 2018, the Court reiterated this line of reasoning in the case F.J.M. v the United Kingdom and argued that the obligation to conduct a proportionality assessment does not automatically apply in cases between two private parties.⁴⁵ The Court argued that two Convention rights are at stake in such cases: the housing right of Article 8 and the property right of Article 1 Protocol No. 1. Private parties often enter into voluntary contractual relationships, such as a landlord–tenant, mortgagee–mortgagor or lender–borrower relationship. These horizontal relationships are governed by national law, and the balance between the interests and rights of both parties has already been determined by national legislation.⁴⁶ Therefore, the Court agreed with the UK Supreme Court that it should not be for domestic courts to alter voluntary contractual relationships based on Article 8 of the Convention.⁴⁷ If a tenant could require a court to conduct a proportionality assessment even if both parties agreed to the terms on which the tenancy would be terminated, this would have a ‘wholly unpredictable and potentially very damaging’ impact on the private rental sector.⁴⁸

Similar to Vrzić, the applicant in F.J.M. voluntarily agreed with eviction as a possible outcome. The applicant entered into an ‘assured shorthold tenancy’, whose purpose is to allow a landlord to retain repossession of the property without the need to give a reason as long as (s)he complies with the notice period.⁴⁹ The nature of this contract drives the Court towards deviating from its previous statement in McCann that ‘any person’ at risk of eviction should be able to have the proportionality of the eviction determined.⁵⁰ The decision of the Court might have been different if the applicant had challenged the nature of shorthold tenancy or the national legislation that is supposed to balance the rights and interests of landlords and tenants.

Some scholars interpret the ruling in F.J.M. as limiting the ‘any person’ statement of the Court in McCann to only those who were evicted by a public authority.⁵¹ According to this interpretation, F.J.M. restricts the possibility of raising an Article 8 defence in private eviction cases.⁵² This reading would be seen as a clear victory for the rights paradigm. However, other scholars propose that the impact of F.J.M. is that ‘any person’ includes anyone who did not agree voluntarily to the terms under which the tenancy would be terminated and the occupants would vacate the home.⁵³ This reading allows for a less severe impact on future private eviction cases as it does not automatically exclude the proportionality test in all horizontal relationships. Such assessment would, for example, still be required in cases that lack contracts and cases where occupants did not enter into contracts voluntarily.

Nevertheless, even if F.J.M. should be interpreted from this more nuanced contractual lens, it still appears to favour the right to property and the presumptive power of creditors and owners. Ownership and the right to property are, by default, dependent on contracts as they establish and recognize the formal transfer and possession of property rights. Without a contract, no legal framework exists to validate or enforce ownership.⁵⁴ In contrast, the concept of home and the right to respect for the home are linked to the ‘factual state of affairs’ and people’s ‘sufficient and continues links’ with their home, which exist independently of any contractual agreements.⁵⁵ The concept of home even transcends the lawfulness of the occupation and could extend to squatters and trespassers.⁵⁶ If a contract, a property law tool which often includes a power imbalance,⁵⁷ could diminish the protection of Article 8, this would indicate a clear property law perspective on housing rights.

This decision has been criticised for contradicting the Court’s previous reasoning and for departing from its own case law without acknowledging this.⁵⁸ Shortly after the publication of F.J.M., Vols predicted that this ruling would probably ‘change the tone and nature of the debate concerning the need of contextual eviction adjudication and the impact of European law on national tenancy law’.⁵⁹ Whether it has affected national law goes beyond the scope of this study, but this article will examine to what extent significant changes could be discovered in the Court’s adjudication of eviction cases since the Vrzić and F.J.M. rulings. I will also examine whether the property law thinking of the Court has had a ripple effect on the eviction of owner-occupiers, leading to a higher chance of winning their cases compared to before the Vrzić and F.J.M. rulings. Additionally, I will assess if homeowners have a better chance of winning compared to non-owners, as they can rely on both Article 8 and Article 1 of the First Protocol. Furthermore, echoing the introduction to this paper, I will not only focus on the procedural safeguards against eviction but also on the more substantive aspects of proportionality to determine the X factor(s) that significantly influences the decision-making of the Court.

3. QUANTITATIVE CASE LAW ANALYSIS

This section is subdivided into five parts and presents the methods and the results of the quantitative analysis of all eviction cases adjudicated by the European Court of Human Rights. Part A discusses the data and methods used for the quantitative analysis. Part B presents the summary statistics of both decisions and judgments, revealing the main characteristics of eviction cases, such as the reason for eviction and the type of tenure. Part C aims to discover tipping points in the adjudication of eviction cases over time, and Part D explores potential X factor(s) influencing case outcomes. The final part zooms in on one particular factor: ownership. As the Court does not issue a separate ruling if an application is deemed admissible and only decides on the merits of the case in judgments, the analyses in Part C, D and E largely focus on judgments (see Part B for an in-depth explanation).

A. Methods

To provide quantitative insights into the relationship between case characteristics and the case outcome, I retrieved all decisions and judgments of the European Court on Article 8 in which the applicant argued that her eviction is a violation of the right to respect for the home as enshrined in Article 8. Together with a team of research assistants, I searched the HUDOC database for English-language cases using the following search terms related to housing and eviction: housing, home, house, eviction, homelessness, evictee, tenant(s), rent, ‘right to adequate housing’, landlord(s) and Article 8. This search yielded 200 relevant decisions (n = 106) and judgments (n = 94) from 1985 up until 2023. We excluded cases that were ‘struck out of the list’ as the Court does not decide on the admissibility nor on the merits of these cases.⁶⁰

We manually collected and coded these cases, ensuring accuracy and consistency through two rounds of coding. We coded the cases for 58 variables, ranging from the date of the case to the Member State involved, the reason for eviction, the type of tenure and the Court’s ruling. Subsequently, I analysed all 200 cases using statistical tests such as the Chi-squared test of independence, Fisher’s exact test and regression analysis. I conducted a series of logistic regressions to estimate the association between the case outcome and various case characteristics, such as the type of tenure, the reason for eviction and the vulnerability of the applicant.⁶¹

Within the presented regression models, the ruling of the Court is called the dependent variable. The variables that I use to explain the dependent variable are known as independent variables or predictor variables. The results of the logistic regression analysis are presented in odds ratios (OR) and standard errors (SE). The odds ratios in this article indicate the positive or negative influence on the probability of the Court finding a violation of Article 8. If the odds ratio is equal to 1, it suggests that there is no association between the independent variable and the case outcome. If the odds ratio of a factor is >1, it indicates that there is a positive association between the variable and the Court finding a violation. If the odds ratio is <1, it indicates that there is a negative association between the variable and the Court finding a violation. The larger the odds ratio, the stronger the association between the case characteristic and the case outcome. The standard error measures the variability or uncertainty in the estimated coefficients of the models. It indicates how much the estimated coefficients might vary if the analysis were repeated on a different sample of data. A smaller standard error suggests that the results are more precise and reliable, while a larger standard error indicates greater uncertainty in the estimates.⁶²

Other results presented in this article are the mean, median (Mdn), the standard deviation (SD) and the p-value. The mean is the average value, the median is the midpoint between the lowest and highest value in a dataset, and the standard deviation provides an indication of how closely the entire set of data aligns with the average value. A larger standard deviation indicates a wider range of values. The p-value, also known as the significance level, presents the probability of two or more distributions being interrelated. The conventional significance level is 0.05. A significance level of 0.05 or lower means that the observed results are very unlikely to have occurred by chance alone. A lower p-value indicates an even greater improbability of chance occurrence.⁶³

B. Eviction Case Law in Numbers

Table 1 provides summary statistics for the entire population of eviction cases adjudicated by the European Court of Human Rights (N = 200) between the first published eviction case on 18 October 1985 up until December 2023. These 200 cases have been lodged against 32 different Member States.⁶⁴ Of these cases, 106 are decisions and 94 are judgments, the two types of rulings the Court issues.

Decisions are rulings of the Court on the admissibility of the case and have the function to filter out cases that do not comply with the procedural requirements (i.e. admissibility criteria), such as failure to exhaust domestic remedies or cases that are so self-evident that the Court deems them ‘manifestly ill-founded’.⁶⁵ In judgments, the Court decides on the merits of the cases and decides whether or not the State violated the Convention. Although the Court does not reach a conclusion on the violation of the Convention in its decisions, these rulings still contribute to the interpretation and application of human rights. Due to the growing caseload, the Court has increasingly put more emphasis on the admissibility criteria in recent years, of which F.J.M. v the United Kingdom is an important example.⁶⁶ Although the Court deemed the application of F.J.M. inadmissible, the Court offered new insights on the interpretation and application of the right to respect for the home in private eviction cases.

The Court classifies each case by level of importance, with the highest being a ‘key case’, followed by levels 1, 2 and 3. Key cases are the most important cases dealt with by the Court. Cases classified as level 1 are of ‘high importance’. These are judgments or decisions with a ‘significant contribution to the development, clarification or modification of its case law, either generally or in relation to a particular State’.⁶⁷ Cases classified as level 2 have medium importance, but nevertheless ‘go beyond merely applying existing case law’.⁶⁸ Cases with a level 3 classification are of low importance as they ‘simply apply existing case law, friendly settlements and strike outs’.⁶⁹ Of all eviction cases adjudicated by the Court, seven are classified as key cases, and 13 are of high importance; all are judgments. The rest of the cases are of medium or low importance. The key and high important judgments are further discussed in the next section on tipping points.

Table 1 also shows the different types of tenure and types of housing that are subject to eviction. Almost half of the cases deal with an owner-occupier that is evicted (48%), and in 86% of the cases the applicants are evicted from either an apartment or a single-household dwelling, which could be a row house or a single detached house. In 81% of cases, the State or a state actor initiates the eviction. In only a small portion of the cases, the eviction is initiated by the landlord, the mortgage lender or the owner not being the landlord. These latter cases concern mostly evictions of roommates or spouses or evictions after someone else inherited the home.

Table 1 also outlines the most common reasons for eviction: demolition, loss of tenancy rights, death of tenancy holder, financial reasons (i.e. arrears) and unlawful transfer of rights. Other reasons for eviction range from anti-social behaviour to absence from the home to expropriation. The category ‘demolition’ refers to demolition orders, owners who already received and used the order and destroyed the house, and the State or state actors who decided to demolish the building, whether for military tactics, land repurposing or new construction. ‘Loss of tenancy rights’ includes cases where the tenancy was based on employment that ended, where the housing was temporary and the lease expired, or where the applicant lost her forced migrant status. The death of a tenancy holder could lead to eviction if the tenancy did not transfer to the residents through succession, meaning that the remaining residents have no occupational rights to stay in the house. Financial reasons include rent and mortgage arrears, as well as failure to pay for services such as plumbing work. The category ‘unlawful transfer of rights’ contains cases about unlawful sales and unlawful housing allocation orders.

In 32.5% of the cases, the applicants mentioned that they are vulnerable. In its case law on Article 8, the Court often emphasises that ‘vulnerable members of society’ should be protected by the State. Vulnerable individuals are, inter alia, children, elderly people, ethnic minorities, people who are physically or mentally ill, and people with a disability.⁷⁰

Table 2 presents the outcome of the decisions and judgments of the Court. In the vast majority of the decisions (86.8%), the Court deemed the application inadmissible. Although this number is reflective of the total amount of applications that the Court declares inadmissible,⁷¹ the numbers in Table 2 present a distorted picture. If an application is declared inadmissible, this decision is final and will be published by the Court. However, the Court generally does not issue a separate ruling if an application is deemed admissible. It advances the case to the judgment phase, in which the Court examines the merits of the complaint. The decision on the admissibility is incorporated into this judgment and is not issued as a separate ruling. Hence, the vast majority of published decisions are decisions in which the Court declares the application inadmissible. The 14 cases in which the Court issued the admissibility decision separately all involve the destruction of property and evictions during the conflict between the Turkish government and the Kurdish population in the 1980s.⁷² These decisions are published because they saw partly on the merits of the case as the Turkish Government contended that the incidents were caused by terrorists for which the government is not responsible. The Court argued that these cases raise ‘complex issues of law and fact under the Convention’, which need a full examination of the merits.⁷³

The judgments numbers shown in Table 2 present a comprehensive overview as the Court publishes all judgments regardless of the outcome. In 72.3% of the judgments, the Court ruled that the State violated the right for respect of the home. This percentage is slightly lower than the percentage of all judgments in which the Court found a violation, which is ~91%.⁷⁴ In the next section, I will explore whether the proportion of violations is consistent over time or if significant shifts (i.e. tipping points) occurred in this pattern.

C. Tipping Points

Echoing the introduction, tipping points could be described as moments in time that brought significant changes to fairly normal or stable phenomena. As Gladwell puts it, a tipping point is ‘that one dramatic moment in an epidemic when everything can change all at once’.⁷⁵ While the tipping point concept has frequently been used to explain changes in beliefs, behaviour, environment and climate,⁷⁶ it could also be used to describe and explain shifts in policies, the interpretation of the law and the adjudication of legal disputes.⁷⁷ In this section, I will study whether there were such tipping points in the adjudication of eviction cases by the European Court. I will explore if there are moments in time when the outcome of the cases significantly altered or ‘tipped’. As explained above, the Court does not issue a separate ruling if an application is deemed admissible. I will, therefore, only focus on changes in the outcome of judgments. However, decisions could still function as tipping points.

Figure 1 presents the judgments of the Court with the outcome of the cases per year (n = 94). Until 2008, the number of human rights violations alternated quite evenly with the number of cases in which the Court rules that there has been no violation of Article 8 (see also Table 3), with an average of 1.2 violations (Mdn = 1.0 SD = 1.6) and an average number of 1.5 no violations found (Mdn = 0.5, SD = 2.0). Nevertheless, the figure shows three interesting spikes until 2008: one in 2001 with five cases in which the Court found no violation, a second one in 2004 where the Court found six violations and a third spike in 2006 with six cases in which the Court found no violation.

The first spike in 2001 is driven by five cases concerning complaints against the United Kingdom brought by five British Roma families (Chapman, Lee, Beard, Smith and Coster) and all adjudicated on the same day (18 January 2001). The applicants bought land and stationed caravans on their land. However, they lacked the required planning permission to do this and were forced to remove the caravans. The Court ruled that the State had a wide margin of appreciation and that there were strong environmental objections to the applicants’ use of their land. The eviction was, therefore, justified, and Article 8 was not violated.⁷⁸ Nevertheless, these cases are classified as judgments with ‘high importance’ as the Court accepted the extension of the scope of Article 8 to include the rights of members of a national minority to have a traditional lifestyle.

Interestingly, one of the six cases that caused the second spike in 2004 is a judgment in which the Court takes a different stance towards the legality of evicting Roma from a caravan site compared to its position in the five cases in 2001. In Connors v the United Kingdom, the Court argued that although the margin of appreciation must be left to the national authorities, the margin tends ‘to be narrower where the right at stake is crucial to the individual’s effective enjoyment of intimate or key rights’.⁷⁹ Moreover, the Court emphasised the lack of procedural protection for the applicant, specifically ‘the requirement to establish proper justification for the serious interference with his rights’.⁸⁰ In contrast to its previous decisions in 2001, the Court now ruled that the State did violate the applicants’ right to respect for the home. While not yet leading to a tipping point in the sense that a significant difference could be noted in the outcome of future eviction cases adjudicated by the Court, the argumentation of the Court in Connors led up the interpretation of the proportionality assessment in McCann 4 years later.⁸¹

The other cases in the 2004 spike are cases in which the applicants complained about the destruction of their property and subsequent forced eviction during the Turkish–Kurdish conflict. In these cases, the Court ruled that the State was responsible for the destroyed property and found this a violation of the right to respect for the home.⁸²

The peak in 2006 is caused by cases about the same conflict, all adjudicated by the Court on the same day (2 February 2006). However, other than in the 2004 cases, the Court finds no violation of Article 8. Although the complaints are fairly similar, the Court was unable to conduct a fact-finding mission like it was able to do so for the 2004 cases.⁸³ Ultimately, due to insufficient evidence and lack of corroboration of the applicants’ claims, the Court concludes there was no violation of the Convention.⁸⁴

Other than the three spikes up until 2006, Figure 1 shows a rapid increase in the number of violations and a decline in the number of judgments in which the Court found no violation since 2008. Table 3 shows that since 2008, the average number of violations spiked to 3.4 (Mdn = 3, SD = 1.7) and the average number of cases in which the court found no violation dropped to 0.5 (Mdn = 0, SD = 0.6).

There was only one eviction judgment in 2008: McCann v the United Kingdom. As the doctrinal analysis in Section 2 already showed, McCann played a pivotal role in the development of the Court’s eviction case law as it established a requirement for national authorities to conduct a proportionality assessment in eviction cases.⁸⁵Figure 1 shows that this effect is also empirically evidenced by the changes in the outcome of eviction cases since the ruling in McCann. A chi-squared test of independence shows that there is a significant relationship between the outcome of judgments and the period of adjudication (i.e. before or after McCann), χ² (1, N = 94) = 19,32, p < 0.001. The p-value of <0.001 means that the probability that the effect of McCann is a coincidence is <0.1%.

Whereas the significance level indicates the importance of individual factors or predictors, more sophisticated statistical tests, such as regression analysis, offer deeper insights into the direction and strength of relationships between factors. A regression model not only confirms McCann’s significance but also quantifies its effect on case outcomes.

Conducting a regression analysis to assess this effect reveals that the odds of the Court finding a violation since McCann are 8.7 times higher compared to the years prior to the ruling (OR = 8,7, SE = 0.5).⁸⁶ These results indicate that the case McCann v the United Kingdom is a tipping point; that ‘one dramatic moment’ when the outcome of eviction cases shifted to significantly more violations of Article 8.⁸⁷

Figure 1 does not indicate any notable shifts prior to or after McCann that significantly altered the outcome of eviction disputes. Although the doctrinal analysis in Section 2 may suggest that Vrzić (2016) and F.J.M. (2018) could be potential tipping points, Table 4 demonstrates that there have been no significant changes in the case outcomes since these two rulings. This does not necessarily imply that these cases had no impact on how eviction cases are adjudicated at European or national level, but rather that, thus far, they have not led to significant changes in case outcomes.

D. X Factors

This part of the article analyses the impact of case characteristics and personal circumstances on the case outcome to study potential X factors that are influential in finding a violation of Article 8. Furthermore, I will measure whether the discovered effect of the McCann case changes if other factors are taken into account. The factors that I focus on are the type of tenure, housing, evictor, the reason for eviction and the vulnerability of the applicants (see Table 5). For this analysis, I solely focus on judgments (n = 94) as the Court does not decide on the merits of the case in decisions. Hence, analysing decisions to measure the impact of case characteristics and personal circumstances would paint a distorted picture.

Table 5 shows the main characteristics of the 94 eviction judgments using the same structure as Table 1, which revealed the summary statistics for the entire population of eviction cases (decisions and judgments). The biggest difference between decisions and judgments in terms of case characteristics is that 23.5% of the entire population concerns public rental housing compared to 40% of the judgments. For all other characteristics, the composition of the sample with judgments is similar to the entire population: the vast majority of cases concern evictions by the State from either an apartment or a single-household dwelling, and demolition and loss of tenancy rights are the most prevalent reasons to evict (see also Table 1).

To assess the impact of these case characteristics on the case outcome and discover potential X factors, I will apply a series of logistic regression analyses. The dependent or outcome variable is the outcome of the case: violation or no violation of Article 8 of the Convention. All analyses measure the chance that the Court finds a violation. The independent or predictor variables are the different case characteristics.

As the sample size is relatively low, I had to regroup some of the variables to prevent overfitting.⁸⁸ Squatting (n = 1) and illegal occupation of land (n = 3) are grouped in the ‘Illegal occupation’ category. The category ‘Other’ within ‘Type of house’ includes university dorms (n = 3) and four cases concerning eviction from a cottage, a hotel that has been used as temporary housing and two previously empty corporate buildings now used as accommodation. The reasons for eviction are rearranged into demolition, contractual issues, (re)claiming possession, arrears and other reasons. Contractual issues include loss of tenancy rights, absence from the home, death of tenancy holder, unlawful transfer of rights and denied planning permission. (Re)claiming possession includes expropriation, new owner, de-nationalisation of property and possession order. The category ‘arrears’ remained the same, and the category ‘other’ includes illegal occupation (n = 4).

Table 6 presents a series of logistic regression models which estimate the chance that the Court finds a violation of Article 8. The results in the table present the coefficients in odds ratio (OR) and the standard error (SE) are placed in parentheses.⁸⁹ The presented odds ratio per category is always in comparison to the reference level, which is the omitted category.⁹⁰ For example, if the odds ratio of a factor is >1, it indicates that the odds of a violation are higher for cases that fall in that category compared to cases in the reference category.

Model 1 presents the influence of the type of tenure on the outcome of the case. It shows that being evicted from a public rental home makes it 10.2 times more likely for the eviction to violate Article 8 compared to when an owner-occupier is being evicted (OR = 10.2, SE = 0.67). This comparison can be reasoned in both directions: ownership cases are 10.2 times more likely to result in the Court finding no violation compared to eviction cases in the public rental sector (OR = 10.2, SE = 0.67). This effect is highly significant (p < 0.001). On the other hand, being evicted from a private rental home does not have a significant impact.

When comparing different types of housing, Model 2 finds that evictions from an apartment have 6.2 times higher odds that the Court finds a violation compared to evictions from a single-household dwelling (OR = 6.2, SE = 0.59). The category ‘other’ shows an unusually high odds ratio and an equally high standard error due to the low number of cases (n = 7), all of which are violations of Article 8. This small number of cases with the same outcome prevents the detection of any reliable and meaningful effects.

Model 3 presents the effects of the different reasons for eviction and shows that the Court is more likely to find a violation if applicants were evicted because of contractual issues and (re)claiming possession by the State or owner of the property compared to evictions due to the destruction of their property (OR = 4.09, SE = 0.55; OR = 11.9, SE = 1.12).

Examining the impact of applicants who mentioned that they are vulnerable, Model 4 reveals that there is no association between this factor and the case outcome (OR = 1.0, SE = 0.48). The different types of evictors, such as landlords and state actors, are left out of the regression analysis because of the high proportion of cases in which the State is the evictor (87%). If there is so little variance in the data, it is not possible to detect any associations.

Now that I have identified which factors have predictive value, I will combine these different variables into the regression analysis. This allows us to assess how the relationships between the identified predictors change, increase predictive accuracy⁹¹ and control for confounding variables.⁹²

Table 6 showed that the strongest relationship exists between the type of tenure and the outcome of the case; more specifically, the odds of the Court finding a violation are 10.2 times higher in evictions from public rental housing compared to owner-occupiers being evicted. Table 7 shows that the magnitude of this effect decreases when controlling for the type of housing (OR = 4.9, SE = 0.77), with the significance level dropping to p = 0.039 (see Model 1). Nevertheless, the model still presents a strong association between public rental and the violation of Article 8. Yet, the effect of the type of housing fails to reach statistical significance (p = 0.087). Further examination reveals a strong overlap between public rent and apartments, with 71.1% of eviction cases in the public rental sector involving evictions from apartments. Therefore, I omit the type of house from subsequent models.

Model 2 demonstrates a similar effect. Where the magnitude of the effect of tenure type increases, the effect of the reason for eviction diminishes. Upon closer examination, a confounding effect is uncovered as 96% of cases about demolition pertain to owner-occupiers. As such, by controlling for the reason for eviction, I inadvertently capture the effect of tenure type twice, leading to distortion of the true associations in the dataset. Consequently, the reason for eviction is also omitted from subsequent models.

To assess whether the effect of the remaining predictor (public rent) persists when controlling for the other identified predictor, the McCann case, I will integrate both variables into a regression analysis. Model 3 reiterates the effect of McCann and shows that cases adjudicated after McCann have a 8.7 times higher odds of resulting in a violation compared to cases adjudicated before the McCann ruling.

In Model 4, both the type of tenure and the McCann case are included in the analysis. The results indicate that while the magnitude of the effect of both factors reduces somewhat, they still emerge as significant predictors of a violation of Article 8. Specifically, public rent and the post-McCann period are associated with higher odds of the Court finding a violation of Article 8.

E. Right to Property

The previous section showed that the odds that the Court finds a violation of Article 8 are significantly higher for public rent eviction cases compared to owner-occupier eviction cases and that, conversely, cases about the eviction of owner-occupiers are 10.2 times more likely to result in the Court finding no violation compared to eviction cases in the public rental sector. This section focuses on this negative impact of ownership on the outcome of eviction cases. I will only focus on cases where homeowners are being evicted, not cases where a private entity evicts tenants, as the dataset only contains 10 such judgments over almost 60 years since the Court’s first eviction ruling. For the same reasons mentioned earlier, the analysis in this Part only includes judgments.

If we not only compare owner-occupiers with public sector tenants but compare owners to non-owners, the magnitude of the effect is slightly lower, but the significance of the influence remains. Figure 2 shows that there is a strong statistically significant difference between the case outcome for owners and non-owners, χ² (1, N = 94) = 16,39, p < 0.0001. A regression analysis reveals that owner-occupiers who are being evicted have a 7.7 times higher odds of losing their case at the European Court than non-owners (OR = 7.7, SE = 0.56).

To examine whether this effect remains consistent over time or if there have been changes in the proportion of violations, particularly since the pro-property rights rulings in Vrzić v Croatia (2016) and F.J.M. v the United Kingdom (2018), Figure 3 presents the temporal trend of case outcomes for cases involving owner-occupiers (n = 45). This trendline shows that the number of cases resulting in violations fluctuates across different years, with an increase in the number of violations since 2016. The proportion of violations changed from 45.5% prior to the Vrzić case to 75% since this ruling. Yet, this difference is not statistically significant.

Reiterating the doctrinal analysis in Section 2, owner-occupiers can not only invoke the protection of Article 8, but can also assert the protection of their property rights under Article 1 Protocol No. 1 of the Convention. To assess if the Court might have ruled that the evictions have violated their right to property instead of their right to respect for the home, I compared these two complaints and the decision of the Court regarding these complaints. Of the 45 cases in which owner-occupiers were evicted, 37 also raised complaints regarding violations of their property rights. Table 8 shows that among these 37 complaints, 14 resulted in a violation, while the Court found no violation in 19 cases, and deemed four complaints regarding Article 1 Protocol No. 1 inadmissible.

Interestingly, in all 14 cases where the Court found a violation of property rights, the Court also found a violation of Article 8. This indicates that the right to property does not serve as a viable alternative or safety net in cases involving owner-occupiers’ eviction. In fact, there were five cases in which the Court found a violation of the right to respect for the home without concomitantly finding a violation of property rights (the Court found no violation or deemed the complaint inadmissible).

4. DISCUSSION AND CONCLUSION

This study is the first to statistically analyse all eviction case law of the European Court of Human Rights. In this final section, I will discuss the identified tipping points that have shifted the trajectory of eviction cases over time. Additionally, I will explore the key factors that significantly influence case outcomes, leading the Court to determine that the applicant has been wrongfully evicted.

The data reveal that the case McCann v the United Kingdom significantly altered the course of the Court’s eviction adjudication. Until 2008, the average number of cases in which the Court found a violation of Article 8 was nearly equal to the number of cases with no violation (1.2 cases per year vs. 1.5 cases per year). However, following the McCann ruling, the average number of violations of Article 8 increased to 3.4 cases per year, while the average number of cases resulting in no violation dropped to 0.5 cases per year. Figure 1 clearly illustrates this divergence in case outcomes: the number of ‘no violation’ cases decreases sharply and remains low, while the number of ‘violation’ cases rises significantly and remains high. In fact, the odds of the Court finding a violation since McCann are 8.7 times higher compared to the years prior to the ruling. This significant shift indicates a pivotal change in the outcome of eviction cases post-McCann, resulting in a sustained disparity in case outcomes thereafter.

An explanation for this effect is the precedent set by McCann, which established a requirement for national authorities to conduct a proportionality assessment in eviction cases. This precedent has created a strong procedural safeguard and lowered the threshold for finding a violation without the need for the Court to conduct an in-depth proportionality assessment and interfere with the margin of appreciation of national authorities. In fact, a deeper analysis of the Court’s judgments in the post-McCann era reveals that the Court found a violation in 42 out of 54 cases that ended up in a breach of Article 8 (77.8%) due to the absence of a proportionality assessment by national authorities, compared to 0 out of 14 cases in which the Court found a violation of Article 8 in the pre-McCann era.

Another explanation for the radical shift is that McCann might have had a filtering effect on the type of cases brought before the European Court. The focus on the proportionality assessment as a mandatory requirement for a justifiable eviction may have caused potential applicants to sustain from bringing their case to the European Court if national authorities conducted a proportionality assessment. As such, McCann could have led to strategic litigation behaviour, which not only impacts the case outcomes but also the type of cases that are brought to the Court.

Nevertheless, McCann represents a clear tipping point that brought significant changes to the Court’s jurisprudence. No other moment in time exhibit a similar notable change. Whereas the push for more contextual adjudication, caused by McCann, was perceived to be diminished in private eviction cases by the rulings in Vrzić and F.J.M., these cases did not alter or shift the outcome of eviction cases. However, the relatively low number of cases involving eviction disputes between two private parties makes it difficult to draw any conclusions on the possible impact of the rulings on such cases.

Similar to McCann, it could be hypothesised that these cases have had a filtering effect, leading to a decrease in number of private eviction cases brought before the Court. Yet, upon closer examination of the composition of cases over time, it becomes apparent that this is not the case. The number of private eviction cases brought before the European Court has not decreased since these two rulings; on the contrary, there were 21 private eviction cases (both judgments and decisions) before the Vrzić ruling (between 1985 and 2016) and 13 cases since Vrzić (between 2016 and 2023). Despite this, the overall rate of private eviction cases is relatively low. This might be explained by the Court’s traditional focus on protecting human rights from state actions. Although eviction orders are typically issued by domestic courts, a state authority, potential applicants might refrain from bringing cases to Strasbourg if the eviction is initiated by a private party, perceiving that the Court is primarily concerned with state-initiated actions.

That the impact of the Vrzić and F.J.M. rulings should not be overestimated is not only suggested by the lack of significant changes in the case outcomes, but is also supported by a more doctrinal analysis. Examining private eviction cases adjudicated since Vrzić and F.J.M. reveals one judgment in which the Court found a violation of Article 8 due to the domestic courts’ failure to conduct a proportionality assessment: Panova and Others v Ukraine.⁹³ In Panova, the applicant was evicted from a residential complex which was formerly State-owned. Yet, the eviction process began 14 years after a private company purchased the complex. While this is reason for the Court to distinguish this case from F.J.M., it does not classify the eviction in Panova as a public eviction. The eviction was ordered by a private owner, indicating that multiple private interests were at stake, not just those of the applicant. Nonetheless, the Court required a proportionality assessment, which illustrates that the Court did not impose a blanket ban on proportionality assessments in private eviction cases.⁹⁴

Interestingly, the Court’s reference to Vrzić in F.J.M. already suggested this more nuanced reading. Although Vrzić v Croatia was the first judgment where the Court ruled that a proportionality assessment is not automatically required in private eviction cases, it still conducted a proportionality assessment in that case. This indicates that the Court did not intend to exclude proportionality assessments from all horizontal cases. Instead, the decisive factor appears to be the clarity of the contractual obligations involved.

In Panova, while the Court recognized the private landlord’s property interests, it ruled that a proportionality assessment was necessary because the terms of the applicants’ occupancy were ambiguous and poorly documented. The Court determined that the eviction would impose ‘an excessive individual burden’ on the applicants.⁹⁵ A similar conclusion was reached by the Court in Brežec v Croatia.⁹⁶ Conversely, in Vrzić and F.J.M., the clarity and voluntary nature of the contractual agreements led the Court to dismiss the Article 8 complaints.

These cases indicate that the Court did not impose a blanket prohibition on proportionality assessments in private eviction cases. Instead, its focus is on whether the terms of the tenancy were voluntarily and clearly agreed upon. This interpretation offers a more nuanced understanding of F.J.M., suggesting that proportionality assessments remain relevant in cases where contractual obligations are less clear or impose an excessive burden on the evictee. This interpretation aligns with the Court’s reasoning in cases preceding Vrzić and F.J.M., where the significance of proportionality was consistently emphasised in various private eviction cases.⁹⁷

As discussed in Section 2, this focus on contractual agreements could still indicate that the Court approaches eviction cases from a property law standpoint. Therefore, this paper also explored the role of ownership and property rights in the Court’s eviction litigation. Interestingly, the data reveal that ownership increases the likelihood of the Court finding no violation of Article 8 in cases involving the eviction of owner-occupiers. Specifically, owner-occupiers have a 7.7 times higher odds of losing their case at the European Court than non-owners. Table 8 shows that this effect cannot be explained by the Court ruling that their right to property has been violated instead of their housing right.

This finding does not support the victory of the rights paradigm, where ownership would typically trump non-ownership. In fact, it is quite the opposite. The analysis shows that non-owners, particularly tenants in the public rental sector, have 5.8 times higher odds of winning their case compared to owner-occupiers, even when controlling for the effect of the McCann case (see Table 7).

This suggests that the Court may prioritize the housing rights of non-owners, potentially due to their socio-economic status or lower income levels compared to property owners. Tenants in public rental housing, who often belong to lower-income groups, appear to receive stronger protection in the Court’s rulings, reflecting its prioritization of socio-economic rights over property rights in the context of evictions.⁹⁸ This is further evidenced by the five judgments in which the Court found that evicting an owner-occupier violated Article 8, but not the right to property (see Table 8). Moreover, the Court tends to grant a narrower margin of appreciation to States when housing rights under Article 8 are at stake, compared to property rights under Article 1 of Protocol No. 1.⁹⁹

As such, this study shows that the European Court does not systematically exclude the proportionality assessment from horizontal relationships nor exhibit bias towards property owners. Rather, the Court emphasizes the importance of voluntary contractual agreements. However, the Court should exercise caution when side-lining human rights in such instances. Contractual relationships often entail power imbalances, and it is the role of human rights law to rectify such disparities.¹⁰⁰

Apart from the McCann ruling and the type of tenure (owner-occupiers and tenants in the public rental sector), no other significant factors have been identified in determining wrongful evictions. In the equation ‘wrongfully evicted = eviction + X’, X equates to the time period since McCann and public rental housing. The relatively small dataset of 200 cases, with only 94 judgments on the merits, limits the ability to statistically analyse the influence of additional case characteristics, such as risk of homelessness, duration of the occupation or a more detailed breakdown of the reasons for eviction or the type of vulnerability, such as age or health status. Therefore, future quantitative research on the Court’s eviction case law should be conducted as more eviction cases will be adjudicated. Additionally, it would be beneficial to study the influence of the Court’s case law on national eviction cases. Yet, while much remains to be explored in future research, this study contributes a new and unique dimension to our understanding of the Court’s eviction case law, highlighting the impact of individual cases and specific case characteristics, and revealing that property rights hold a supporting rather than a central role in eviction cases.

ACKNOWLEDGEMENTS

This research was completed as part of the EVICT project (www.eviction.eu) funded by the European Research Council Research Grant under grant agreement No 949316.

Footnotes

Author notes