Human Rights Law Review, 2023, 23, 1–18

https://doi.org/10.1093/hrlr/ngad013
Article

The Indivisibility of Human Rights: An

Empirical Analysis

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Jan Essink∗ , Alberto Quintavalla∗∗ and Jeroen Temperman∗∗∗
ABSTR ACT
This article aims to test whether human rights have an indivisible nature. To do that, we perform
correlation analysis and Granger causality tests to test 1) the relationship within socio-economic rights
and 2) between socio-economic rights and civil-political rights. The results show that certain socio-
economic rights have mutual reinforcing relationships, lending support to the existence of widespread
indivisibility. This finding yields relevant policy implications. Given their financial constraints, states
could make use of the existence of widespread indivisibility, in combination with the progressive
implementation clause, to foster the efficient allocation of resources for human rights implementation.
Furthermore, this article shows that the intensity of indivisibility varies depending on the income
category of states: the indivisible nature of socio-economic rights is more intense in low-income
countries while seems to achieve a saturation point at the highest levels of human rights compliance.
We, thus, propose to define this phenomenon as ‘indivisibility saturation’. Lastly, our findings detect
a more complex picture for the indivisibility principle between the two classes of human rights. While
widespread indivisibility does not follow from the tests, important unidirectional relationships between
different human rights exist and are equally important for human rights policy-making purposes.
KE Y WO R DS: civil-political rights, Granger causality tests, human rights implementation
indivisibility principle, socio-economic rights

1. INTRODUCTION
The human rights record of many countries in the world is not as rosy as one may hope
for.1 Naturally, international (human rights) institutions, the international community of states
and NGOs and civil society organisations continue calling abusive regimes to account for
and improve their human rights records by way of respecting, protecting and fulfilling the
obligations these very same states have committed to. However, this objective is not always
easily attainable due to financial and political reasons. Some governments suffer from a lack of
economic resources; others are simply not willing to respect human rights. Moreover, while both
* Erasmus University Rotterdam, The Netherlands. E-mail: [email protected].
** Erasmus University Rotterdam, The Netherlands. E-mail: [email protected].
*** Erasmus University Rotterdam, The Netherlands. E-mail: [email protected].
1 See, e.g. Amnesty International, Amnesty International 2021/2022 Report. The State of World’s Human Rights (Amnesty
International, 2022) reporting globally increasing inequalities and Human Rights Watch, World Report 2022. Events of 2021
(Human Rights Watch, 2021) expressing concerns for autocracies.

Received: March 23, 2022. Revised: November 15, 2022. Accepted: April 25, 2023
© The Author(s) [2023]. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution NonCommercial-NoDerivs
licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution
of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited.
For commercial re-use, please contact [email protected]
2 • Human Rights Law Review

at the international- and domestic-level strategies for the implementation of human rights law
are typically in place, there is no guidance on such pertinent policy questions as to what measures
to prioritise, from the perspective of optimal human rights promotion, nor any feed-back loop
on the short- or longer-term effects of the measures taken compared to alternative options that
were open. This entails that although human rights law promotes all measures improving the
human rights situation, what human rights policy is optimal right now and within the current
context or constraints at hand remains a black box.
Thereby, providing states with concrete guidance for human rights implementation could

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

represent a possible improvement of the status quo. Countries can in fact benefit from exploring
what types of relationships exist among human rights. For instance, knowing that investing on
a certain right may have positive spillover effects on another right, but not vice versa, could
lead (politically willing) governments to prioritise the implementation of the former over the
latter. But how would it be possible to devise clear and well-defined strategies for human rights
implementation in practice? And, most importantly, would international human rights law
provide concrete guidance for human rights implementation?
Although this type of questions is of interest to virtually every country in the world, finding
a clear answer is a tall order. The reason lies in the tendency from various United Nations
bodies to shy away from framing human rights discussions in terms of hierarchy or priorities.
In fact, the international community frequently refers to human rights as ‘universal, indivisible
and interdependent and interrelated’.2 These essential characteristics are inherent to all human
rights.3 For example, the Committee on Economic, Social and Cultural Rights opens General
Comment No 21 on the right of everyone to take part in cultural life by stating that ‘[c]ultural
rights are an integral part of human rights and, like other rights, are universal, indivisible and
interdependent’.4 Likewise, the Committee Against Torture deems the obligations to prevent
torture and other cruel, inhuman or degrading treatment or punishment to be ‘indivisible,
interdependent and interrelated’.5 Similarly, the Committee on the Elimination of Discrimina-
tion against Women declared that women’s right to a life free from gender-based violence is
‘indivisible from and interdependent with other human rights’.6
This approach makes it difficult to derive well-defined guidelines on how to advance human
rights in each country when there are budgetary constraints. The theoretical argument is that
all human rights are universal and indivisible and, accordingly, a prioritisation is not possible.
Concurrently, the endorsement of various United Nations bodies for the universal and indivisi-
ble nature of human rights, thus, facilitates the perception that these human rights’ features are
undisputed in international human rights law. However, at a closer view, the human rights law
scholarship has occasionally challenged these features.
In fact, the interrelatedness of human rights may resemble more an aspirational statement
since one must concede that, in practice, not all human rights support each other. For instance,
Ashford shows that it is well possible to secure the right to freedom from torture irrespective of

2 United Nations, Alternative Approaches and Ways and Means within the United Nations System for Improving the Effective
Enjoyment of Human Rights and Fundamental Freedoms. UN Doc. A/34/RES/48 (1977); United Nations, Declaration
on the Right to Development. UNGA Res. 41/128, 4 December 1986; United Nations, Vienna Declaration and Programme
of Action, World Conference on Human Rights. UN Doc. A/CONF.157/24 (1993).
3 Donnelly and Whelan, International Human Rights (2017) at 57; Whelan, Indivisible Human Rights. A History (2010) at 1.
4 United Nations, Committee on Economic, Social and Cultural Rights, General Comment No 21 on the right of everyone to
take part in cultural life (art. 15), E/C.12/GC/21 (2009).
5 United Nations, Committee Against Torture, General Comment No 2 on Implementation of article 2 by States parties,
CAT/C/GC/2 (2008), see however Ashford, ‘The alleged dichotomy between positive and negative rights and duties’ in
Beitz and Godin (eds), Global Basic Rights (2009) at 167–178.
6 United Nations, Convention on the Elimination of all Forms of Discrimination Against Women, General recommendation
No. 35 on gender-based violence against women, updating general recommendation No. 19, CEDAW/C/GC/35 (2017).
 The Indivisibility of Human Rights • 3

whether the right to subsistence is secured.7 This latter consideration makes the indivisibility
principle even more difficult to accept. This is because existing literature tends to consider
the indivisibility principle as the strongest type of supportive relationship in the tripartite
formulation of indivisible, interdependent and interrelated.8
Admittedly, the discussion on the existence of the indivisibility principle has attracted con-
siderable interest from a theoretical viewpoint. Most human rights scholars tend to acknowl-
edge the indivisibility of human rights but debate its intensity within the human rights law
framework.9 In simpler terms, while there is consensus that the full implementation of a human

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

right requires the achievement of other human rights, scholars discuss whether these supporting
relationships exist between all rights or only some of them—something that can be defined as
‘the indivisibility hypothesis’. In the last couple of decades, quantitative studies have begun to
complement the theoretical discussion on the indivisibility hypothesis.10 This empirical strand
of scholarship has not yet provided conclusive evidence on whether this hypothesis tends to
hold. Additionally, these studies have so far solely focused on political and civil rights due to the
paucity of data on socio-economic rights. In other words, the indivisibility principle has been
tested only with regard to what type of relationships exist among civil and political rights. As
Soiffer and Rowlands put it, the class of civil and political rights has become ‘a proxy measure
for human rights in general’.11
Hence, this article aims to test the ‘indivisibility hypothesis’ by examining the relationships
between socio-economic rights themselves and, subsequently, between socio-economic rights
and civil-political rights. The importance of this work is twofold. First, this analysis would
enrich the existing literature by testing whether and, if so, what type of indivisibility principle
holds. The ‘indivisibility hypothesis’ will thus be tested in the context of a single class of human
rights (i.e. socio-economic rights) as well as between two classes (i.e. socio-economic and civil-
political). Second, it would yield important policy implications. The findings could in fact
show whether a state investing resources on the implementation of a given human right may
cause positive spillover effects on other human rights. This last consideration is even more
important in the realm of socio-economic rights due to the dose of discretion left to states
through the progressive implementation clause (Article 2(1) ICESCR). Contextualising the
work, the results would possibly show how to render the efforts of the governments that want to
protect and promote human rights more effective.
This article is structured as follows. Section 2 discusses the indivisibility principle in human
rights law. It shows that the principle is enshrined in various human rights law instruments and
that this legal construct is, for most scholars, not only an aspirational claim, but an actual legal-
normative principle that should be respected. While Section 3 introduces the data, Section 4
outlines the method employed to test empirically the ‘indivisibility hypothesis’ and runs the
analysis. In Section 5, we discuss the results and their possible policy implications. We then
conclude.

7 Ashford supra n 10.

8 Nickel, ‘Rethinking Indivisibility: Towards a Theory of Supporting Relations between Human Rights’ (2008) 30 Human
Rights Quarterly 984, at 987–991; Donnelly and Whelan supra n 8 at 60; Whelan supra n 8 at 6.
9 Nickel supra n 14; Gilabert, ‘The Importance of Linkage Arguments for the Theory and Practice of Human Rights. A
Response to James Nickel’ (2008) 32 Human Rights Quarterly 425.
10 Milner, Poe and Leblang, ‘Security Rights, Subsistence Rights and Liberties: A Theoretical Survey of the Empirical
Landscape’ (1999) 49 Human Rights Quarterly 403; Minkler and Sweeney, ‘On the Indivisibility and Interdependence
of Basic Rights in Developing Countries’, (2011) 33 Human Rights Quarterly 351; Soiffer and Rowlands, ‘Examining the
Indivisibility of Human Rights: A Statistical Analysis’ (2018) 17 Journal of Human Rights 89.
11 Soiffer and Rowlands supra n 16 at 90.
4 • Human Rights Law Review

2. THE INDIVISIBILITY PRINCIPLE

A. The Human Rights Law Instruments and the Legal Theory Literature:
Brothers in Faith
The indivisibility principle is one of the key tenets of international human rights law.12 This
principle was not included in the Universal Declaration of Human Rights but emerged only
subsequently. Specifically, the first reference to the indivisible nature of human rights can be
found in the 1968 Proclamation of Teheran.13 Its paragraph 13 affirms that human rights are
indivisible insofar as the ‘full realization of civil and political rights without the enjoyment of

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

economic, social and cultural rights is impossible’.14 The focus of this proclamation was not on
the indivisibility of all human rights taken together, but rather on the distinction between civil
and political rights, on the one hand, and economic, social and cultural rights, on the other.
The principle of indivisibility, thus, seems to have emerged to address the political divide
between two different classes of human rights. This human rights dichotomy had characterised
the international human rights law structure since its beginning. For instance, the initial draft of
the Covenant on Human Rights was extended to encompass socio-economic rights. The failure
of this proposal subsequently led to the adoption of two distinct international human rights
law treaties in 1966: the International Covenant on Civil and Political Rights (ICCPR) and the
International Covenant on Economic, Social and Cultural Rights (ICESCR). It is in this context
that two years later the Proclamation of Teheran introduced the indivisibility principle.15 The
international human rights law community needed, so the argument runs, an instrument to
ensure a solid and harmonious implementation of both classes of human rights.16
Further human rights law documents have continued to reflect this view. For example, the
1977 General Assembly Resolution 32/130 states that ‘[a]ll human rights and fundamental
freedoms are indivisible and interdependent; equal attention and urgent consideration should
be given to the implementation, promotion, and protection of both civil and political, and
economic, social, and cultural rights’.17 Likewise, the Declaration on the Right to Development
and the General Comment No 22 on the right to sexual and reproductive health stresses the
indivisible nature between such rights and the other civil, political, economic, social and cultural
rights.18 Interpreting the indivisibility principle as a rhetorical tool to lessen the distinction
between two classes of human rights made its way in regional treaties, too. The African Charter
on Human and Peoples Rights adopts a similar approach when stating that ‘the satisfaction
of economic, social and cultural rights is a guarantee for the enjoyment of civil and political
rights’.19
The indivisibility principle has acquired additional meanings over time.20 For instance,
the indivisible nature of human rights started becoming an aspirational claim to facilitate
the adoption of a holistic view when it comes to the human rights system.21 Post-Cold War

12 See supra n 8.
13 United Nations, Proclamation of Teheran: Final Act of the International on Human Rights, UN Doc.
A/CONF.32/41(1968).
14 ibid.
15 See, respectively, International Covenant on Civil; and Political Rights 1996, 999 UNTS 171 and International Covenant
on Economic, Social and Cultural Rights 1996, 993 UNTS 3.
16 Whelan supra n 8 at 7–8.
17 UN General Assembly, Alternative approaches and ways and means within the United Nations system for improving the
effective enjoyment of human rights and fundamental freedoms, General Assembly Resolution 32/130 (1977) at para 1(a).
18 United Nations, Committee on Economic, Social and Cultural Rights, General Comment No 22 on the right to sexual
and reproductive health (art. 12), E/C.12/GC/22 (2016) at para 10; see also United Nations, Vienna Declaration and
Programme of Action, World Conference on Human Rights, UN Doc. A/CONF.157/24 (1986) at para 6.
19 Organization of African Unity, African Charter on Human and Peoples Rights, OAU Doc. CAB/LEG/67/3 rev. 5, 21 ILM
58 (1982).
20 Whelan supra n 8 at 7; Donnelly supra n 8 at 60.
21 Donnelly supra n 8 at 60–61.
 The Indivisibility of Human Rights • 5

instruments considerably fostered that conceptual shift of emphasis, notably, the 1993 Vienna
Declaration, which refers to all human rights as, inter alia, indivisible.22 Further developments
in international human rights law contributed to expanding the meaning of the indivisibility
principle. Specifically, certain human rights treaties such as the Convention on the Elimination
of All Forms of Discrimination against Women and the Convention on the Rights of the
Child have weakened the distinction between the two classes of rights (by way of including
specimen from both categories indiscriminately) and, relatedly, contributed to provide new
lenses through which to look at the indivisibility principle.23 The same goes for such regional

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

efforts at human rights codification as the EU’s Charter of Fundamental Rights of the European
Union, comprising both civil-political ‘freedoms’ and what it refers to as ‘solidarity’ rights (socio-
economic rights), among other rights.24 The process of ‘mainstreaming’ human rights has
further weakened distinctions between categories of rights. For instance, while participatory
rights are explicitly enshrined in the ICCPR, the Committee on Economic, Social and Cultural
Rights has frequently ‘read’ participatory rights ‘into’ socio-economic rights by way of the lens
of due diligence obligations.25 Accordingly, such rights as the right to housing, food, water,
employment, all could be breached under circumstances by dint of failures in the participation
department. Such mainstreaming exercises, in the absence of an explicit treaty basis, have been
called ‘the representation par excellence of the doctrine of the indivisibility of human rights’,
which is indivisibility in action.26
At the same time, the literature notes that this shift in the conceptualisation of indivisibility
has given way to enhance the ambiguity of its meaning to the point of becoming a ‘meaningless
sloganeering’.27 This criticism notwithstanding, the United Nations has never elaborated on the
specific meaning of the indivisibility principle.
The indivisibility principle has then attracted considerable interest among scholars given
its obscure definition and the far-fetched implications that would derive therefrom.28 The
definitional issue pertaining to its meaning is in fact not just a philosophical inquiry, but a
relevant question closely associated with its implementation.29 Considering human rights as
possessing an indivisible nature complicates the theoretical and practical discourse on whether
the implementation of certain human rights may be preferred over others.30 This discussion
acquires considerable relevance especially when the United Nations adopts human rights law
instruments that may call for a prioritisation of human rights. A striking example is the pro-
gressive implementation clause of the ICESCR according to which states are called to realise
economic, social and cultural rights progressively and based on the actual resources at their
disposal.31 Further instruments on a differentiated implementation are the Limburg Principles

22 See supra n 7.
23 de Beco, ‘The indivisibility of Human Rights in Light of the Convention on the Rights of Persons with Disabilities’ (2019)
68 International Comparative Law Quarterly 141.
24 EU Charter of Fundamental Rights of the European Union (2000), proclaimed by the EU on 7 December 2000, subse-
quently incorporated into the Treaty of Lisbon (effective 1 December 2009).
25 Temperman, ‘Public Participation in Times of Privatization: A Human Rights Analysis’ (2011) 4 Erasmus Law Review 43 at
60–66.
26 ibid. at 61.
27 Scott, ‘Reaching Beyond (Without Abandoning) the Category of Economic, Social, and Cultural Rights’ (1999) 21 Human
Rights Quarterly 633, at 644.
28 Nickel supra n 14; Nickel, ‘Indivisibility and Linkage Arguments: A Reply to Gilabert’, (2010) Human Rights Quarterly 439;
Gilabert supra n 15. In addition, one should note that discussing the actual meaning of indivisibility had the consequence
of shifting the debate from its historical role on the divide between the two classes of human rights to the human rights law
structure as a whole: Scott supra n 33 at 644.
29 Quintavalla and Heine, ‘Priorities and Human Rights’, (2019) 23 International Journal of Human Rights 679, at 681.
30 Shue, Basic Rights: Subsistence, Affluence and US Foreign Policy (1996) at 26; Scott supra n 33 at 644–645; Quane, ‘A Further
Dimension to the Interdependence and Indivisibility of Human Rights? Recent Developments Concerning the Rights of
Indigenous People’, (2012) 25 Harvard Human Rights Journal 49 at 50.
31 Article 2(1) International Covenant on Economic, Social and Cultural Rights supra n 21.
6 • Human Rights Law Review

(nos 25 and 72) and the Maastricht Guidelines (no. 10): by establishing that certain subsistence
rights should be respected regardless of the actual availability of resources, these instruments
implicitly acknowledge that once the minimum core obligations are fulfilled, states enjoy a
certain dose of discretion in terms of human rights implementation.32
Hence, scholars began to advance certain hypotheses on what indivisibility means in human
rights law. Most academics agree that the indivisibility of human rights entails a mutually rein-
forcing relationship among different human rights.33 As quickly mentioned in the introduction,
the main theoretical disagreement lies in the intensity of these interactions. It is usually contested

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

whether indivisibility should apply to all human rights (i.e. so-called ‘system-wide indivisibility’)
or only to a limited number of human rights (i.e. ‘widespread indivisibility’).34 This latter
disagreement, however, does not impede us to reach a common definition of indivisibility.
Hence, in line with previous research, this article will refer to the ‘indivisibility hypothesis’
as the existence of bidirectional supporting relationships between (all or some) human
rights.35
B. Empirical Literature
The debate on the intensity of the indivisibility principle within the human rights law structure
has rarely led scholars to theoretically contest the indivisibility principle as such. The main
theoretical scepticism to indivisibility pertains to the challenges that could arise by considering
socio-economic rights in the same way as political and civil rights.36 Instead, disputing the actual
existence of an indivisible nature of human rights usually arises if scholars adopt an empirical
viewpoint. As De Beco put it, ‘[t]he indivisibility of human rights would be an exaggeration from
an empirical point of view’.37 According to Minkler and Sweeney, there are two main strands of
empirical literature on states’ provision of human rights.38 The first strand of literature examines
the determinants of states’ human rights records.39 In this regard, the literature emphasises, not
surprisingly, a state’s ability and a state’s willingness to provide human rights to its citizens. As
to a state’s ability to provide human rights, several scholars find that more developed countries
have a better human rights record than less developed countries.40 As to a state’s willingness
to provide human rights, it is reported that democracies usually have a better track record than
non-democracies.41 These works, thus, included a series of variables that represent predictors
for the respect of both civil-political (e.g. democracy) and socio-economic (e.g. Gross Domestic
Product per capita) rights.

32 Alston and Quinn, ‘The Nature and Scope of State Parties’ Obligations under the International Covenant on Economic,
Social and Cultural Rights’ in Ssenyonjo (ed), Economic, Social and Cultural Rights (1987) 3–76.
33 Nickel supra n 14; Gilabert supra n 15 at 432–434.
34 Nickel supra n 14 at 988; Gilabert supra n 15 at 430. A related discussion pertains to what is the specific type of supporting
relationships among human rights. According to Nickel, these relationships should be of a very strong type: one human right
is ‘indispensable or necessary’ to another human right. Another view, instead, is to reduce the strength of such relationships
by defining them as ‘highly useful’.
35 Nickel supra n 34 at 441; Soiffer and Rowlands supra n 16 at 93.
36 Roth, ‘Defending Economic Social and Cultural Rights: Practical Issues Faced by an International Human Rights
Organization’ (2004) 21 Human Rights Quarterly 63 at 63; Neier, ‘Social and Economic Rights: A Critique’, (2008) 13
Human Rights Quarterly 1 at 1–3.
37 de Beco supra n 29 at 148.
38 Minkler and Sweeney supra n 16 at 354–364.
39 ibid.
40 See, e.g. McCormick and Mitchell, ‘Human Rights Violations, Umbrella Concepts, and Empirical Analysis’, (1988) 49(4)
World Politics 476; Henderson, ‘Conditions Affecting the Use of Political Repression’, (1991) 35(1) Journal of Conflict
Resolution 120 at 134; Poe and Tate, ‘Repression of Human Rights to Personal Integrity in the 1980s: A Global Analysis’,
(1994) 88 American Political Science Review 853 at 866–867.
41 See, e.g. Henderson, ibid. at 132–3; Poe and Tate, ibid. at 866; Davenport and Armstrong, ‘Democracy and the Violation of
Human Rights: A Statistical Analysis from 1976 to 1996’ (2004) 48(3) American Journal of Political Science 538 at 551.
 The Indivisibility of Human Rights • 7

What is more is that these empirical models have occasionally considered the very same
relationship between (indicators that could represent) civil-political and socio-economic rights.
For instance, previous scholarship showed that economic development is inducive to greater
protection of subsistence rights.42 Likewise, it has been found that the expansion of political
liberties leads to a higher income per capita.43 This body of knowledge typically did not frame
that discussion in terms of indivisibility, but rather in terms of growth v. equity.44
There is a second strand of literature aiming to analyse explicitly the interrelation of states’
records under the different types of human rights.45 Naturally, this second strand is the most

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

relevant for any discussion of the ‘indivisibility hypothesis’. The literature that empirically
examines the interrelatedness of states’ human rights records dates back from the late 1990s.
Milner and co-authors examine four different categories of human rights (security rights,
subsistence rights, political liberties and economic freedom).46 By constructing indices of
these different categories of human rights and performing correlation analysis, they find that
generally the different categories of human rights tend to be achieved together.47 More recently,
Minkler and Sweeney similarly find that states tend to secure subsistence and security rights
simultaneously.48
These articles provide a good first step to evaluate whether different categories of human
rights have an indivisible nature, but there is still room for refinement. A first point of criticism to
these studies comes from Soiffer and Rowlands: a strong correlation between two human rights
cannot establish bidirectionality, at most showing simultaneity.49 To overcome this limitation,
Soiffer and Rowlands propose to run the Granger causality tests and show evidence for a
bidirectional relationship between some, but not all political and civil human rights.50 Neatly
summing up the empirical state of the art (until 2017), Soiffer and Rowlands show that most
statistical studies typically focusing on various civil and political rights establish significant
unidirectional relationships. Indivisibility’s key features of bilateralism or multilateralism are not
corroborated, nor is the respective, say, pace right for indivisibility. For true indivisibility one
would expect a joint climax, but simultaneous achievement is rarely established in these studies.
The second point of criticism we raised, instead, relates to the fact that these studies do not focus
on socio-economic rights or, when they do, their index does not take into consideration the
principle of progressive realisation.51 Thus, the added value of our article is that it incorporates
both criticisms in testing the existence of the ‘indivisibility hypothesis’. In other words, this
article performs both correlational analysis and the Granger causality tests within a single class
of human rights (i.e. socio-economic rights) as well as between two different classes (i.e. socio-
economic and civil-political) by using an index that incorporates the principle of progressive
realisation.

42 See, e.g. Park, ‘Correlates of Human Rights: Global Tendencies’ (1987) 9 Human Rights Quarterly 405; Pritchard, ‘Human
Rights and Development: Theory and Data’ in Forsythe (ed), Human Rights and Development (Palgrave Macmillan, 1989)
329; Mitchell and McCormick, ‘Economic and Political Explanations of Human Rights Violations’ (1988) 40 World Politics
476; Poe and Tate supra n 47.
43 Kaufman, ‘Human Rights and Governance: The Empirical Challenge’ in Alston and Robinson (eds), Human Rights and
Development: Towards Mutual Reinforcement (OUP, 2005) 352.
44 See, e.g. Moon and Dixon, ‘Basic Needs and Growth-Welfare Trade-offs’, (1992) 36 International Studies Quarterly 191.
45 Minkler and Sweeney find that at the time of writing there were few studies that examined the interrelation between different
types of human rights. The most notable study at the time was Milner et al. supra n 16.
46 Milner et al. supra n 16 at 438.
47 ibid.
48 Minkler and Sweeney supra n 16 at 365–367.
49 Soiffer and Rowlands supra n 16 at 93.
50 ibid.
51 Milner et al. supra n 16 at 365–367.
8 • Human Rights Law Review

3. DATA
The human rights domain is characterised by a paucity of data measuring the implementation of
human rights in the various states due to the difficulty to divide human rights into plausible and
measurable constituent components. The limited data availability has constrained the ability
of researchers to carry out empirical studies in the past.52 It is only more recently that this
situation has changed with the introduction of some datasets measuring the implementation
of a certain set of human rights.53 The two most well-known data sources in the field are the
Freedom in the World Dataset and the Cingranelli and Richards Human Rights Dataset.54 Both

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

datasets measure for a large number of countries the extent to which a state complies with certain
human rights over a relatively long timeframe. However, the focus of these datasets is exclusively
on political and civil human rights. This implies that empirical works relying on these datasets
cannot test the indivisible nature of economic, social and cultural rights.
Given that the first part of this article is devoted to evaluating to which extent socio-economic
rights are indivisible among themselves, we use the Index of Social and Economic Rights
Fulfilment (the SERF index) by the Economic and Social Rights Empowerment Initiative at
the University of Connecticut.55 The SERF index provides the scores for five socio-economic
human rights (1) the right to education; (2) the right to health; (3) the right to housing; (4)
the right to food; and (5) the right to work for 221 different countries between 2007 and
2017. Arguably, this is the only relatively comprehensive dataset that provides information on
the extent to which states manage to meet their economic and social rights obligations under
human rights law. Importantly, the dataset is developed to take into consideration the concept
of progressive realisation, as established in Article 2(1) of the ICESCR. The scoring that each
country receives from the SERF index then incorporates the extent to which states could ensure
the realisation of socio-economic rights based on the maximum of their own available resources.
The SERF index manages to do that by using an ‘Achievement Possibilities Frontier’ (APF).
The APF frontier shows to which extent states have the capacity to provide the economic and
social rights to their citizens based on their own GDP per capita. Figure 156 provides an example
of the APF with regard to the right to education: the richer the countries in terms of GDP per
capita, the more they must avoid children effectively being deprived of education. The index
calculates how far a state finds itself from the APF (see, for instance, the red, orange, and yellow
lines for Kenya, Sudan, and India, respectively). The closer a country finds itself to the APF, the
higher its score on the SERF index. If a country finds itself on the APF, then it has the full score
of 100.
Although the SERF index offers us the possibility to analyse the extent to which the afore-
mentioned five different socio-economic rights are indivisible, its use is not without limitations.
First, the dataset is considerably shorter compared to other existing datasets on human rights.
For example, while the CIRI dataset runs between 1981 and 2011, the SERF index only runs
between 2007 and 2017. This limitation is not fatal, even though our preference would in
principle be for a dataset with a longer timeframe. The reason is that for answering the question
of whether socio-economic rights are indivisible, we need to analyse to which extent socio-
economic rights mutually influence each other over time. For this purpose, datasets with longer

52 McCormick and Mitchell supra n 45 at 510.

53 Milner et al. supra n 16.
54 See, respectively, Freedom in the World, Freedom House, available at: https://freedomhouse.org/report/freedom-world and
Cingranelli, Richards and Clay, CIRI Human Rights Data Project, available at: http://www.humanrightsdata.com/p/data-
documentation.html [last accessed 18 March 2022].
55 Index of Social and Economic Rights Fulfillment, The Economic and Social Right Empowerment Initiative, https://serfindex.u
conn.edu/ [last accessed 18 March 2022].
56 ibid.
 The Indivisibility of Human Rights • 9

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Fig. 1. Achievement possibility frontier for “Percentage of children not stunted”.

timespans would be ideal.57 Second, the dataset is unbalanced: some countries lack data on
socio-economic rights in some years. For instance, in the time span 2007–2017, Austria has
missing data as to the obligation to provide health services to its citizens for the years 2007
and 2008. Third, the SERF index distinguishes only between high-income and low-income
countries. While the following analysis will maintain this distinction to have more granular
observations for these two subsets of countries, we readily acknowledge that a broader number
of subsets (i.e. high income; middle income and low income) would have depicted the states’
socio-economic configurations more accurately. Despite these three deficits, the SERF index
is a good data source for the article’s aims. Not only does this dataset measure socio-economic
rights, it also takes into account the principle of progressive realisation. Additionally, the fact that
the SERF index dataset covers a large number of countries mitigates the problem of a relatively
short time span.
The second aim of the paper, instead, is to test whether the ‘indivisibility hypothesis’ holds
between two classes of human rights: socio-economic rights and civil-political rights. Hence, the
data used for civil-political rights are from Freedom House. This dataset covers a large number
of countries between 2003 and 2021 by accruing a score for compliance with political rights
(ranging from 0 to 40) and civic liberties (ranging from 0 to 60).58 This dataset is useful for our
purpose because in addition to covering many of the same countries as the SERF index, it also
covers the same time range.
In fact, the other alternative to the Freedom House’s civil and political rights indexes (i.e.
the CIRI Human Rights Dataset) has a very limited overlap in data coverage with the SERF
index. Specifically, the SERF dataset and the CIRI dataset have an overlap period of four years

57 See infra the part detailing the Granger causality tests.

58 Freedom House calculates these scores based on 10 types of political rights and 15 types of civil liberties. Experts allocate
between 0 to 4 points to every right. The maximum score for political rights equals therefore to 40 and for civil liberties to
60. See Freedom House supra n 57.
10 • Human Rights Law Review

only (i.e. from 2007 to 2011), while the SERF dataset and the Freedom House dataset have a
ten-year overlap period (i.e. from 2007 to 2017). This implies that opting for using the indices
from the CIRI dataset instead of the indices from Freedom House would significantly reduce
the number of datapoints to perform the analysis, thereby reducing the reliability of the analysis.
Nonetheless, in the Annex, we performed the same analysis with the CIRI’s physical integrity
rights and empowerment indices to extend the analysis to another set of predictors for the
protection of civil and political rights.59
After having collected this data, we standardised these scores such that they hold values

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

between 0 and 100. In this way, both the used scores of the socio-economic rights and the scores
of political-civil rights range between 0 and 100.

4. ANALYSIS
A. Introductory Remarks
In this section, we perform the analysis to evaluate whether the indivisibility principle holds for
socio-economic human rights as well as between socio-economic and civil-political rights. In
fact, as previously mentioned, indivisibility among civil and political rights is more well-charted
territory, the two types of dynamics at hand are not charted at all and our aim hence is to fill
these two specific gaps. For this purpose, and as anticipated, we define indivisibility as a bilateral
supporting relationship among human rights. In other words, two human rights are indivisible if
the respect for a first human right increases the respect for another human right, and vice versa.
This definition is in line with the one given by previous research.60 This section will thus be split
into two main subsections: the empirical analysis on the indivisibility hypothesis between socio-
economic rights (subsection B) and between the two classes of human rights (subsection C).
Based on this definition, the first part of this article investigates the type of relationship
existing between all the possible pairs of human rights included in the dataset for socio-
economic rights (i.e. the right to education, the right to health, the right to housing, the right to
work, the right to food). The methodology employed in this paper statistically tests the bilateral
relationships within the pairs of human rights. First, we use correlation analysis to verify whether
there is correlation within the pairs. The main limitation of a correlation analysis is that it does
not yield any finding about causation, which is whether an increase in implementation of human
right A positively affects the implementation of right B. This first part of the statistical analysis
serves therefore merely to get a rough idea of whether potential relationships within the pairs
of human rights exist. To address the issue of causation, instead, we use the Granger Causality
tests in the second part of our statistical analysis. The Granger Causality tests would allow
us to identify what type of relationships between human rights pairs exist (i.e. bidirectional,
unidirectional or no relationship).
The results stemming from our analysis will show if (1) there are bidirectional relationships
between all pairs of human rights considered (i.e. systemwide indivisibility); (2) there are
bidirectional relationships between some pairs of human rights (i.e. widespread indivisibility);

59 The political rights index of Freedom House is based on an evaluation of the electoral process, the degree of political
pluralism and participation and the general functioning of government while the civic liberties index is based on an
evaluation of the freedom of expression and belief, associational and organisational rights, the rule of law, personal autonomy
and individual rights. The physical integrity rights index of the CIRI dataset is constructed by combining indicators on
torture, extrajudicial killing, political imprisonment and citizens’ disappearance, while the empowerment index is based on
combining indicators on the degree of movement rights, freedom of speech, freedom of assembly and association, workers’
rights, electoral self-determination and freedom of religion.
60 While Nickel claims ‘two rights are indivisible as the claim that they are mutually indispensable (strongly supportive
in both directions’, Soiffer and Rowlands state ‘following working definition of indivisibility: the strong, simultaneous,
and bidirectional relationship between either all (as is the case for systemwide indivisibility) or some (as is the case for
widespread or partial indivisibility) human rights’. See, respectively, Nickel supra n 34 at 441 and Soiffer and Rowlands
supra n 16 at 93.
 The Indivisibility of Human Rights • 11

Table 1. Classification of strength of correlation coefficients

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

and (3) there are no bidirectional relationships between human rights (i.e. no indivisibility). We
shall now turn to the description of the correlation analysis and the Granger causality tests. The
second part of this article then continues by examining the relationship between these individual
socio-economic rights and the aggregate scores measuring the extent to which (a) political rights
and (b) civil liberties are offered.61 For this purpose, we again employ a correlation analysis and
Granger Causality tests.
B. The Indivisibility Hypothesis between Socio-Economic Rights
(i) Correlation analysis
With the purpose to get some initial information about the strength between the different pairs
of human rights, we perform a simple correlation analysis. Since the dataset consists of five
different human rights (i.e. the right to education; the right to health; the right to housing; the
right to food and the right to work), the analysis will involve ten distinctive pairs of human
rights. This involves the pairwise calculation of Pearson correlation coefficients of these ten
pairs of social and economic rights, as provided by the SERF database. The Pearson correlation
coefficient shows the bivariate correlation between two variables. The coefficient takes a value
between −1 and 1. If the value is negative, then there is a negative correlation between the two
variables while the opposite is true if the value is positive. The further the coefficients deviate
from 0, the stronger the correlation. In Table 1, we provide a classification of Pearson correlation
coefficients.
If economic and social rights would be indivisible one would thus expect correlation coef-
ficients that are positive and not close to zero. In Tables 2 and 3, we present the outcomes of
the Pearson correlation coefficients between the ten pairs of economic and social rights for,
respectively, the subset of low-income and high-income countries in the dataset.
Table 2 shows that the correlation within the different pairs of socio-economic human rights
for the subset of low-income countries can be classified as either moderate (5 pairs) or strong
(5 pairs). For example, while the right to food has a positive moderate correlation with all the
other rights, the right to housing is strongly correlated with the right to health.
Table 3 shows that the correlation within the different pairs of socio-economic human rights
for the subset of high-income countries is generally weaker as compared to the subset of low-
income countries. The strength of the correlations can be classified as either strong (3 pairs),
moderate (4 pairs), weak (2 pairs), or very weak (1 pair). The only two pairs for which the
correlation is stronger for the subset of high-income countries, as compared to the subset of
low-income countries, are Food-Health and Work-Food.

61 According to Freedom House, the criteria to rate political rights are a) electoral process; b) political pluralism and
participation; c) functioning of government and d) personal autonomy and individual rights while the ones to rate civil
liberties are a) freedom of expression and belief; b) associational and organizational rights; c) rule of law and d) personal
autonomy and individual rights. See Freedom House supra n 57.
12 • Human Rights Law Review

Table 2. Overview of correlation coefficients between the different categories of socio-economic

rights (sample: low-income countries)

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Table 3. Overview of correlation coefficients between the different categories of socio-economic
rights (sample: high-income countries)

In sum, based on this correlation analysis it is more likely that we will find supportive evidence
for the indivisibility principle in the subset of low-income countries than in the subset of high-
income countries.62 Granger causality tests are used to examine this.

5. GRANGER CAUSALITY TESTS

The existence of indivisibility among human rights entails that the respect for a first human right
has positive spillover effects on the respect for a second human right, and vice versa. In other
words, the relationship between human rights should be bidirectional. A simple correlation
analysis presents some limitations to verify whether such a relationship exists in fact. The reason
is twofold. First, a strong correlation between two variables (X and Y) indicates that there is
a relationship between variables without shedding light on the nature of this relationship. For
example, it could even be that the human rights have no relationship due to the existence of a
contravening omitted variable c that would influence both variables X and Y. In this case, the
relationship would be defined as spurious. Second, even if we assume that there is an actual
relationship between X and Y, a simple correlation analysis would not show whether it is X to
cause Y, or Y to cause X. It might for instance be that human right X simply follows human right Y,
while human right Y does not follow human right X. In other words, a high correlation between
two human rights would not necessarily imply that there is a bidirectional relationship between
the two human rights, only pointing at the existence of a unidirectional relationship. For these

62 The existence of a moderate or strong correlation among all the human rights investigated might appear supportive of the
indivisibility hypothesis. In fact, this is only a tentative conclusion that should be substantiated by Granger causality tests.
 The Indivisibility of Human Rights • 13

two reasons, a high correlation score in the respect for two human rights does not allow us to
draw definitive conclusions on whether these two human rights are indivisible.
Hence, there is a need to complement the correlation analysis with a series of Granger
causality tests. As Soiffer and Rowlands argue, this technique mitigates both problems associated
with a simple correlation analysis.63 Granger causality is an approach pioneered by Granger that
is frequently used in social sciences to identify the causal relationship between two variables.64
To understand the reasoning behind the Granger causality test, let us imagine two variables
X and Y. To see whether X causes Y, the Granger Causality performs a regression analysis by

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

regressing variable Y on (a K number of) lagged values of Y itself and a (L number of) lagged
values of variable X. If one of the L lagged variables of X is statistically significant, then the test
finds that passed values of variable X have explanatory power for the current value of variable Y.
In this case, the test finds that variable X Granger causes variable Y.
The Granger causality tests would suggest that human rights X and Y are indivisible if
human right X causes human right Y, and human right Y causes human right X. This is called
bidirectional causality. It requires, thus, two separate Granger causality tests to verify whether
two human rights are indivisible. Admittedly, Granger causality tests were originally designed for
time series data (data for one cross-section units over time). In the last three decades, however,
Granger causality tests have become popular in the context of panel data (data for multiple cross-
section units over time), too. This latter development presents a significant advantage since it
increases the number of datapoints. The implication is that it is also possible to run these tests
for dataset with a shorter time dimension.
Furthermore, Hood et alia identify two important inferential issues for the Granger causality
tests in a panel data setting due to heterogeneity between the cross-section units, which in our
case are states.65 The first one relates to the omitted variable bias. The implication is that if
we do not control for this type of omitted variables, the Granger causality test may find that
the provision of human right X causes the provision human right Y even if that is not the case
in practice. The reason is that there might be an underlying factor, the so-called confounder,
that influences the provision of both human right X and the provision of human right Y.66 We
follow previous literature and include fixed effects into the Granger causality tests to mitigate the
influence of potential confounders.67 The second problem can be that different cross-section
units have heterogeneous causal relationships between two variables. It might for instance be
the case that for a developing country such as Benin there is no causal relationship between the
provision of human right X and human right Y while for a developed country such as Germany
there is. This is a direct consequence of pooling a set of very heterogeneous countries together
in one dataset. To limit this, we divide our analysis into low-income and high-income countries.
In Tables 4 and 5, we show the results of the analysis for, respectively, the subset of low-income
and high-income countries.

63 Soiffer and Rowlands supra n 16 at 97.

64 Granger, ‘Investigating Causal Relations by Econometric Models and Cross-spectral Methods’ (1969) 32 Econometrica 424;
Hood, Kidd and Morris, ‘Two Sides of the Same Coin? Employing Granger Causality Tests in a Time Series Cross-section
Framework’ (2008) 16 Political Analysis 324.
65 Hood et al. supra n 66 at 326.
66 The literature finds (see supra n 45) for instance that democracies are more likely to provide human rights to their citizens.
Imagine that in fact there is no causal relationship between variable X and Y but that countries that have better developed
democratic institutions are more likely to provide both human right X and Y of a better quality. Some countries will thus
provide both human rights X and Y of a better quality, but this is purely due to the fact that they have better functioning
democratic institutions. Failing to take this influence of the quality of democratic institutions into account means that we
might observe that there is a link between the provision of human right X and the provision of human right Y even if this is
not the case.
67 Hood et al. supra n 66 at 326; Soiffer and Rowlands supra n 16 at 97.
14 • Human Rights Law Review

Table 4. Overview of the relationships between the different pairs of socio-economic human rights
(sample: low-income countries)
Bidirectional relationship Unidirectional relationship No relationship
Education—Housing Education—Food Housing—Work
Health—Housing Education—Work
Health—Food Health—Work
Food—Housing Health—Education
Work—Food

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Note: for the unidirectional relationship, the first mentioned right causes the second mentioned right

Table 5. Overview of the relationships between the different pairs of socio-economic human rights
(sample: high-income countries)
Bidirectional relationship Unidirectional relationship No relationship
Food—Housing (Housing—Food: negative) Health—Education (negative) Education—Housing
Housing—Work Education—Food (negative) Education—Work
Work—Health Health—Housing
Health—Food
Work—Food
Note: for the unidirectional relationship, the first mentioned right causes the second mentioned right

The Granger causality tests indicate that for the subset of low-income countries there is
a bi-directional relationship between five pairs of human rights, a unidirectional relationship
between four pairs (where the first mentioned right causes the second mentioned right yet not
vice versa) and no relationship between one pair.
The Granger causality tests show for the subset of high-income countries a bi-directional
relationship between two pairs of human rights, a unidirectional relationship between three
pairs and no relationship between five pairs. It indicates, consistently with the earlier executed
correlation analysis, that there is generally less evidence for indivisibility for the subset of high-
income countries than for the subset of low-income countries. Three pairs have a bi-directional
relationship for the subset of low-income countries which evaporates for the subset of high-
income countries (i.e. Education-Housing, Health-Housing and Health-Food). Similarly, two
pairs that have a unilateral relationship for the subset of low-income countries turn into no
relationship for the subset of high-income countries (i.e. Education-Work and Work-Food).
There are, however, two pairs for which a new relationship is established. One pair turns the
unidirectional relationship around (i.e. for the subset of low-income countries Health causes
Work while for the subset of high-income countries this is the other way around) and a second
pair turns from no relationship to a bi-directional relationship (i.e. Housing-Work). For the
remaining three pairs, it is interesting to note how the type of relationship becomes negative
(Housing-Food, Health-Education and Education-Food): the implementation of the first men-
tioned human right causes negative spillover effects on the second mentioned human right. For
instance, in the bilateral relationship of Food-Housing, while the right to food positively impacts
the right to housing, the latter negatively influences the former.
In line with previous studies, these tests reject the existence of a systemwide indivisibility for
the subset of both low-income countries and high-income countries.68 Systemwide indivisibility
would have required the existence of bidirectional relationships between all pairs of human

68 Milner et al. Supra n 16; Soiffer and Rowlands supra n 16.

 The Indivisibility of Human Rights • 15

Table 6. Overview of correlation coefficients between socio-economic rights and civil-political rights
(sample: low-income countries)

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Table 7. Overview of correlation coefficients between socio-economic rights and civil-political rights
(sample: high-income countries)

rights, let alone the absence of negative relationships among pairs of human rights. However, for
both subsets of countries, we find some evidence for bidirectional relationships (and positive in
both ways). Following the classification in the introduction of section 4, we find thus support
for widespread indivisibility.
A. The Indivisibility between the Two Classes of Human Rights
(i) Correlation analysis
The correlation analysis shows that for the subset of low-income countries there is either a very
weak (4 times) or a weak degree (6 times) of correlation between the different types of socio-
economic rights and political-civil rights (see Table 6).
Interestingly, the degree of correlation between the different types of socio-economic rights
and political-civil rights is stronger in all cases for the subset of high-income countries as
compared to the subset of low-income countries. The evidence shows for the former subset
three times a strong correlation, 2 times a neutral correlation and 5 times a weak correlation
(see Table 7).69
(ii) Granger causality tests
To get a better idea of the nature of the correlation, we again perform Granger causality tests.
The Granger causality tests performed for the subset of both low-income and high-income
countries do not indicate the existence of any bidirectional relationships between the different
types of socio-economic rights and both political rights and civil liberties (see Tables 8 and 9,
respectively). The tests therefore reject the indivisibility hypothesis between the two classes
of human rights. It is, however, worth mentioning that a limited number of unidirectional
relationships exist. For instance, enhanced respect for the right to housing decreases the respect
for both political rights and civil liberties in low-income countries. On the other hand, a better
compliance with the right to work increases the respect for political rights, only.

69 As evidenced in tables 1 and 2 of the Annex, performing the same analysis with the CIRI’s physical rights and empowerment
indices shows generally for both the subset of low- and high-income countries weaker relationships. A potential explanation
for this is the lower number of datapoints.
16 • Human Rights Law Review

Table 8. Overview of the relationships between socio-economic rights and civil-political rights
(sample: low-income countries)
Bidirectional Unidirectional relationship No relationship
relationship
Housing—Political Rights (Negative) Political Rights—Education
Work—Political Rights Political Rights—Health
Housing—Civil Liberties (Negative) Political Rights—Food
Civil Liberties—Education

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

Civil Liberties—Health
Civil Liberties—Food
Civil Liberties—Work
Note: for the unidirectional relationship, the first mentioned right causes the second mentioned right

Table 9. Overview of the relationships between socio-economic rights and civil-political rights
(sample: high-income countries)
Bidirectional Unidirectional relationship No relationship
relationship
Civic Liberties—Education (Negative) Political Rights—Education
Work—Civic Liberties Political Rights—Health
Political Rights—Housing
Political Rights—Food
Political Rights—Work
Civic Liberties—Health
Civic Liberties—Housing
Civic Liberties—Food
Note: for the unidirectional relationship, the first mentioned right causes the second mentioned right

Turning to the subset of high-income countries, it is possible to observe that only two
unilateral relationships exist. Namely, a better record for civil liberties would have negative
spillover effects into the area of respect for the right to education and a better record in the area
of the right to work influences positively respect for civil liberties.70

6. DISCUSSION
The findings of the correlation analysis and the Granger causality tests show that the indivisible
nature is not ubiquitous in the human rights framework. On the other hand, it is possible to argue
that widespread indivisibility characterises the human rights framework due to the existence of
mutual reinforcing relationships between many socio-economic human rights. This implies that
the indivisibility principle is not merely a rhetorical tool, but sometimes also finds expression
empirically. Interestingly, the result of this study on socio-economic rights is consistent with
previous studies on political and civil rights.71 A more complex picture is given instead for
the indivisibility principle between the two classes of human rights. While it is not possible to
argue for the existence of widespread indivisibility, some important unidirectional relationships

70 Tables 3 and 4 of the Annex provide the results of the analysis with the CIRI’s physical rights and empowerment indices.
These results point at a limited support for the indivisibility hypothesis. For example, in the sub-sample of high-income
countries, there is a bidirectional relationship between the physical integrity index and the right to work, a unidirectional
relationship between the right to education and the physical integrity index and no significant relationship in the rest of the
cases.
71 Soiffer and Rowlands supra n 16.
 The Indivisibility of Human Rights • 17

between different human rights exist. These findings do not undermine indivisibility per se,
neither as a doctrine nor as an aspirational principle. Depending on the context, political rights
may have the ‘major role’ toward ‘the solution of economic privation’ as Sen predicted.72 But
universal human rights enjoyment at the same time remains a multi-fronted battle and no one
set of rights forms a magic wand. In the absence of system or even widespread indivisibility, it
in fact is precisely the specific dynamics, including the unidirectional impacts observed in this
paper, that may prove pertinent to policy-making.
Another interesting finding is that the degree of widespread indivisibility among socio-

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

economic human rights changes depending on the income-category of states, low-income versus
high-income countries. Specifically, the indivisibility of socio-economic rights in the subset of
high-income countries is less intense than in the subset of low-income countries also due to
the existence a larger number of negative relationships. This finding tends to contradict part
of previous research according to which less intense supporting relationships between human
rights characterise developing countries.73 In fact, there may be different levels of human rights
implementation—each entailing a different strength in the supporting relations between human
rights.74 Since developing countries focus solely on a low level of human rights implementation,
there would be no possibility to have—so the argument runs—strong indispensable relations
between rights.75
Nonetheless, our results show that more intense supporting relationships among socio-
economic rights are more likely to exist in low-income countries, which are the ones having more
significant budgetary restraints. Plausible theoretical explanations may include the following.
Developed countries enjoy a more mature political and economic environment that push them
for fulfilling relatively high human rights standards. In this way, the implementation of a human
right at a high level would not likely contribute to realise another human right at its high level of
implementation. In other words, less practical spillover effects from one right to another exist
once a certain level of human rights implementation is reached. An example may clarify this
point. The right to water requires that each individual has access to a certain amount of water.
While the amount of water would be 50 l/d at the low level of implementation, the amount of
water would be 150 l/d at the high level of implementation.76 It is reasonable to expect that
the marginal utility of having the first 50 l/d will certainly be higher as compared to the last
100 l/d. Accordingly, the supporting relationship between the right to water and another right
(say, the right to health) will likely be more intense at a low level of implementation. This latter
consideration does not mean that the linkage arguments are of little use at a high level of human
rights implementation. It simply means that linkage arguments at a high level of implementation
become more a relevant rhetorical tool to ensure that no human right is left behind. Relatedly,
one may propose to talk in terms of ‘indivisibility saturation’: mutually reinforcing spillover
effects appear to occur in a more widespread fashion at the level of basic yet advancing human
rights enjoyment, while somewhat of an indivisibility saturation point seems to be reached at
the highest levels of human rights compliance.
In addition, the scope of this work goes beyond the mere act of testing the ‘indivisibility
hypothesis’. Finding mutual reinforcing interactions between certain human rights may have
relevant implications for the policy approaches adopted by states, too.77 According to the

72 Sen, ‘Freedoms and Needs’ in The New Republic 32 (1994).

73 Nickel supra n 14; Nickel supra n 34.
74 Nickel supra n 34 at 444. According to previous research, the differing levels of implementation may concern individuals,
groups, or regions. See Nickel supra n 14 at 993.
75 Nickel supra n 14 at 997–999.
76 Gleick, ‘The Human Right to Water’ (1998) 1(5) Water Policy 487 at 496.
77 Gilabert supra n 15.
18 • Human Rights Law Review

existing human rights law framework, states are in principle called to implement all human rights
altogether. However, Article 2(1) ICESCR introduces a sort of limitation to this obligation.78
As mentioned above, in the context of economic, social and cultural rights, states enjoy a
certain margin of discretion in human rights implementation based on the availability of their
resources.79 This progressive implementation clause occupies even more centre stage for the
specific case of developing countries facing considerable budget constraints.
Hence, identifying what are the human rights that have a mutual reinforcing relationship
could serve as a tentative ‘roadmap’ for states and their policies on how to steer resources for

Downloaded from https://academic.oup.com/hrlr/article/23/3/ngad013/7187934 by guest on 16 August 2023

human rights implementation. For instance, our analysis shows that the right to housing has
a bidirectional relationship with all the other human rights except for the right to work. A
state facing budget constraints that would like to allocate resources to improve the respect for
the socio-economic rights considered in this study should not channel all the funding to the
improvement of the right to housing. This is because there would be no positive spillover effects
on the right to work. Likewise, if the state had to decide on whether to invest some remaining
funds on either the right to education or the right to food, it should prefer the former given the
unilateral direction from education to food.
To sum up, the contribution of this study in testing ‘the indivisibility hypothesis’ is twofold.
First, it provides empirical evidence that there is widespread indivisibility among socio-
economic rights whereas only unidirectional relationships between different classes of human
rights exist. Second, the specific findings can provide governments with clear and well-defined
guidelines on how to structure a differentiated human rights implementation when it comes to
allocating funds.

7. CONCLUSION
Every country in the world should respect and promote human rights. An imposing and lofty
challenge, this objective requires time to be achieved. In fact, most of the countries have a hard
time in complying with all human rights obligations. Financial and political problems account
for this situation. A possible way to improve the human rights record may be to provide states
with concrete guidance for human rights implementation. Specifically, it would be helpful to be
informed of what types of relationship exist among human rights. This problem is not moot in
contemporary discussions on human rights implementation: this information would contribute
to water down both financial and political problems. On the one hand, the allocation of resources
should privilege the specific human right having most positive spillover effects on other human
rights. On the other hand, human rights organisations and NGOs could decide to channel most
of their energies into targeted actions. Namely, they could pressure the states that consistently
violate several human rights on the respect of the specific human right(s) which would have the
biggest spillover effects on the other rights.
It is against this background that this article analysed the relationships within socio-economic
rights and between socio-economic rights and civil-political rights. The analysis showed that
certain human rights have mutual reinforcing relationships whereas others have (positive and
negative) unidirectional or no relationship. We believe that these results are worth highlighting.
At aggregate level, they show that the indivisible nature of human rights is not always ever-
present and the commitment of the human rights law framework to it is aspirational. At a more
granular level, the findings can become of use to several countries who are willing to advance
their human rights situation more effectively.

78 Article 2(1) International Covenant on Economic, Social and Cultural Rights supra n 21.
79 ibid.