Katarina Schwarz, University of Nottingham
Prohibiting Slavery, Criminalising Enslavement
The Failures of the International R e gi me Governing Human Exploitation
For years, international lawyers and scholars considered the definition of slavery to encapsulate the
status of de jure slavery and nothing more. With the outpouring of international instruments from the
1800’s to 1926 when the Slavery Convention was signed, international law was seen to have brought
about the gradual, absolute prohibition of slavery and the slave trade. With the rise of the ‘white slave
trade’, human trafficking, and ‘modern slavery’, the displacement of the orthodox interpretation by
Allain’s study of the 1926 Convention’s travaux préparatoires, and figures produced by Walkfree, the
ILO and the IOM suggesting that 40.3 million people lived in conditions of ‘modern slavery’ in 2017,
greater interrogation of the international regime governing human exploitation is overdue.
This paper draws upon a comprehensive review of domestic legislation concerning slavery, servitude,
institutions and practices similar to slavery, forced labour, and human trafficking of all 193 UN Member
States in comparison to their existing international obligations.
It explores trends in the criminalisation
of human exploitation, displacing the presumption of campaigners and scholars that slavery is illegal
in every country. Failures of the international regime are thereby brought to light, inviting
consideration of the future of international law governing human exploitation.
For decades the prevailing opinion that slavery could only exist where legal ownership was
present shrouded the reality of a large-scale failure on the part of States. A failure to enact
domestically one of the most fundamental prohibitions known to international law – the prohibition
Although the rise of ‘human trafficking’ in international and domestic law gave the
This paper is based upon a continuing study by Jean Allain and Katarina Schwarz, initiated in 2016 by a Working
Group comprised of Jean Allain (chair), Amanda Kramer, Aisling Ledwith, Katarina Schwarz, and Egle
Vasiliauskaite. Consideration of domestic legislation is drawn from open, publicly available sources including:
Constitute Project (www.constituteproject.org), Slavery in Domestic Jurisdictions (www.qub.ac.uk/slavery), and
UNODC SHERLOC (www.unodc.org/cld/v3/sherloc/) databases; the Engen Report, UN Doc. E/2673, 1955; and
the Awad Report, UN Doc. E/4168/Rev.1, 1966.
In Barcelona Traction, the International Court of Justice recognised the prohibition against slavery as a jus
cogens norm entailing obligations erga omnes. Case concerning the Barcelona Traction, Light and Power
Company Ltd (Belgium v Spain)  ICJ Rep 3, 32 at .
prohibitions against slavery, servitude, and forced labour a new place in international law, and gave
rise to a new wave of domestic legislation, it also concealed the lacuna in domestic implementation
measures. This implementation gap leaves slavery, institutions and practices similar to slavery,
servitude, and forced labour outside the boundaries of national legislation in many States, except to
the extent that they form a subsidiary element of an offence of trafficking in persons. In light of the
international legal frameworks governing these forms of exploitation, this ought to be considered an
unacceptable situation for international legal scholars.
The collection of domestic legislation concerning slavery, servitude, forced labour, and human
trafficking shows that national engagement with international law governing human exploitation has
been erratic, irregular and incomplete. This paper not only displaces the notion that slavery is now
effectively abolished in all States, but reveals the extent to which states have neglected to do so,
highlights trends in the provisions outlawing human exploitation, and underscores the value of
empirical research in assessing the implementation of international law more generally.
The research project and legislative database
In order to assess the extent to which the identified forms of human exploitation have been
prohibited in domestic law, the constitutional, criminal and labour legislation of 193 UN Member
States was compiled, and provisions dealing with slavery, servitude, institutions and practices similar
to slavery, forced labour and human trafficking were drawn from these texts. These provisions were
selected not only for direct references to the relevant forms of human exploitation, but by their
relation to some element of the prohibition in question. Thus, a provision dealing with buying and
selling human beings would be classified as relevant to slavery, regardless of whether the term
‘slavery’ was used in the text. Likewise, provisions addressing just one of the institutions and practices
similar to slavery identified in the 1956 Convention would be deemed relevant, even if it only
constituted a partial prohibition of that form of exploitation. In total, over 700 pieces of domestic
legislation were analysed, with thousands of relevant provisions drawn from their texts.
Obligations undertaken under international instruments in these areas, including through
regional arrangements, were also incorporated into the dataset. Relevant international obligations
with regard to slavery, servitude, forced labour, and human trafficking were drawn from seven core
international instruments. Working definitions of the different forms of exploitation are drawn from
these texts – with the definition of slavery found in the 1926 Slavery Convention, Servitude in the 1956
Supplementary Convention on the Abolition of Slavery, forced or compulsory labour in the 1930
Forced Labour Convention, and human trafficking in the 2000 Protocol to the Transnational Organised
Crime Convention concerning Trafficking in Persons. For the purpose of this paper, ‘servitude’ and
‘institutions and practices similar to slavery’ are both dealt with under the broader category of
servitude in order to reflect the substance of the prohibition against servitude in general international
The outputs from this research include a dataset of international obligations and domestic
legislation concerning human exploitation; and individual reports for each State summarising the
domestic law in place, identifying the membership status of that State to the relevant treaties, and
drawing a conclusion as to compliance with, or breach of, international obligations in force. The former
allows for quantitative analysis of domestic implementation incorporating a range of additional
variables (including membership in a range of regional organisations and geographic location), while
the latter enabled qualitative analysis of the domestic legislation enacted, and the degree to which
this meets international obligations and trends in national provisions. This paper explores the key
findings of this research including an overview of the implementation gap by numbers and by
Although the quality of the dataset is somewhat limited by the requirement that legislation
be publicly available in an online format, the existence of the 1966 Awad and 1955 Engen Reports
significantly offset the difficulty of locating more obscure legislation.
Searches were conducted in
See Jean Allain, Slavery in International Law: Of Human Exploitation and Trafficking (Martinus Nijhoff 2013)
Engen Report, UN Doc. E/2673, 1955; Awad Report, UN Doc. E/4168/Rev.1, 1966.
multiple languages in order to reduce language barriers, and relevant provisions translated in order
to allow for universal analysis. Even absent a strong degree of certainty, this research provides a tool
for understanding the extent and scope of domestic implementation, advocating for the passage of
legal provisions in all States, and encouraging the satisfaction of the reporting obligations contained
in article 7 of the 1926 Slavery Convention
and article 8 of the 1956 Supplementary Convention.
research may also be seen as a roadmap for research into domestic implementation of international
obligations in a range of different areas of international law, highlighting the value of empirical
research in assessing the role and effectiveness of international law.
As a jus cogens norm of international law, the prohibition against slavery applies to all states
irrespective of their membership in any particular treaty. Yet, despite the universality of the
customary prohibition, only thirteen States do not have treaty obligations concerning slavery under
the 1926 Slavery Convention 1956 Supplementary Convention, or the 1966 ICCPR (see fig. 1).
A similarly low number of States do not have treaty obligations in respect of the remaining
forms of exploitation, with only seventeen States free from treaty obligations concerning servitude,
Article 7, as amended by the 1953 Protocol requires parties communicate laws and regulations adopted with
a view to application of the Convention to the Secretary-General of the United Nations. Slavery Convention
(adopted 25 September 1926, entered into force 9 March 1927) 60 LNTS 253.
Article 8 imposes a similar obligation on parties to communicate laws, regulations and administrative measures
to the Secretary-General and additionally requires the Secretary-General to communicate this information to
the Economic and Social Council. Supplementary Convention on the Abolition of Slavery, the Slave Trade, and
Institutions and Practices Similar to Slavery (adopted 7 September 1956, entered into force 30 April 1957) 266
Fig 1. In t er n at io n al Obl i ga t io ns to P r oh ib i t H uma n Ex plo i t at io n
fourteen without obligations with regard to forced labour, and twenty-three without obligations
under the Palermo Protocol concerning trafficking in persons. In total, only six States globally are not
party to any of the key treaties concerning the four types of human exploitation addressed in this
It is therefore clear that the vast majority of States have treaty-based obligations in relation
to human exploitation, and the need to rely upon jus cogens status is limited.
Domestic implementation by numbers
Despite the widespread existence of obligations concerning human exploitation, and often of
specific obligations of criminalisation,
the enactment of legislation prohibiting human exploitation is
variable. The reality of domestic implementation is starkly (if somewhat simplistically) captured by the
number of States who have adopted some form of legislation dealing with human exploitation. In
particular, the extent to which human trafficking has dominated contemporary discourse on
exploitation in international law is revealed both by the frequency of domestic legislation (see fig. 2)
and the dates of adoption of this legislation. Although these figures cannot speak to the effectiveness
of implementation measures, they indicate the degree to which the first step in combating human
exploitation domestically – the enactment of legislative prohibition – has been taken.
By far the most promising example of an international prohibition in domestic law, human
trafficking finds an almost universal place in national legislation. In fact, more States have such
provisions than have ratified the Palermo Protocol.
The trend towards criminalising trafficking in
Bhutan, Brunei-Darussalam, the Marshall Islands, Palau, Tonga and Tuvalu are not party to the 1926 or 1956
slavery conventions, the 1930 or 1957 forced labour conventions, the 1966 ICCPR or the 2000 Palermo Protocol,
although the Marshall Islands are party to the 1998 Rome Statute of the ICC covering enslavement.
Articles 3, 5, and 6 of the 1956 Supplementary Convention (above n 6) entail obligations to criminalise slavery,
the slave trade and mutilation. Article 5 of the Palermo Protocol requires the criminalisation of trafficking in
persons – Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children,
Supplementing the United Nations Convention against Transnational Organized Crime (adopted 15 November
2000, entered into force 25 December 2003) 2237 UNTS 319 (Palermo Protocol). Article 6 of the 1926 Slavery
Convention (above n 5) relatedly requires that severe penalties be imposed for infractions of laws giving effect
to the purposes of the Convention. Article 25 of the 1930 Forced Labour Convention requires that illegal exaction
of forced or compulsory labour be made a penal offence – Forced Labour Convention, C29 (adopted 28 June
1930, entered into force 1 May 1932) 39 UNTS 55.
While 170 States have ratified the Palermo Protocol on Trafficking in Persons, 186 have domestic legislative
provisions dealing with human trafficking.
persons gained traction in recent decades, with more than half of all States having adopted specialist
In defining trafficking, however, many States chose not to adopt the specific definition
outlined in the Palermo Protocol (or enacted unrelated provisions prior to its adoption). A handful of
States, for instance, have thus far failed to criminalise internal trafficking, prohibiting only crimes
involving transnational movement. This may be connected to the Palermo Protocol’s attachment to
the Convention against Transnational Organised Crime, and the related presumption that trafficking
is situated within the arena of organised crime, and takes place across international borders.
on sexual and child exploitation, and the omission of labour exploitation within provisions related to
trafficking in persons is an even more significant trend – often including the specific inclusion of
women and children as potential victims to the exclusion of adult males. This runs counter to the
Palermo definition which equally recognises victims of trafficking irrespective of gender, despite a
stated priority for women and children evidenced by the full title.
Yet, despite considerable failures
in the definition of trafficking in international law and in the accompanying provisions required by
international law, this form of human exploitation has seen by far the most comprehensive
criminalisation in domestic law.
Over 100 States have adopted independent, specialist anti-trafficking legislation.
See Jean Allain, above n 3, 353.
Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children.
81 63 104102 60 81 18658
Slavery Servitude Forced Labour Human Trafficking
Fig 2. Domestic Legislation concerning Human Exploitation
Legislation No Legislation Constitutional provisions Criminal Provisions Labour Provisions
Painting a less positive picture of domestic implementation, and despite 180 States having
treaty based obligations to prohibit slavery, and all States having such obligations under customary
international law, only 140 States were found to have legislative provisions in some way prohibiting
slavery. Of these 140 States, 81 had provisions in their Constitutions while 102 had enacted criminal
provisions. Given that the 1956 Supplementary Convention specifically requires the criminalisation of
slavery and the slave trade, this relatively low level of domestic enactment leaves many States in
breach of their obligations.
Moreover, this figure includes those States with only partial prohibitions
of slavery and the slave trade, insufficient to meet the international definition. In light of the growing
recognition of the relevance of ‘slavery’ as a form of exploitation which continues to exist in legal
this represents a significant implementation gap which ought to be the focus of future
advocacy and research.
Perhaps unsurprisingly given the confusion over the substance of ‘servitude’ in international
law, the criminalisation of servitude is the least implemented domestic measure within the forms of
human exploitation studied. Just over half of States (101) have some form of legislative prohibition
against either servitude as a distinct concept or any of the institutions and practices similar to slavery
enumerated in article 1 of the 1956 Supplementary Convention. Of these States, 63 have
constitutional provisions and 60 criminal sanctions. Given that right to be free from servitude (as well
as slavery and forced labour) is incorporated into the ICCPR as a non-derogable right to which 168
States are party,
and that the 123 parties to the 1956 Supplementary Convention undertook to ‘take
all practicable and necessary legislative and other measures’ to ensure abolition in a now concluded
transitional period, the failure of 92 States to prohibit even one of the forms of servitude represents
a significant failure of the international regimes.
See above n 8.
See for instance Caso Trabajadores de la Hacienda Brasil Verde v Brasil (Judgement) Inter-American Court of
Human Rights Series C No 318 (20 October 2016) and The Queen v Tang  HCA 39.
Articles 4(2) and 8, International Covenant on Civil and Political Rights (adopted 16 December 1966, entered
into force 23 March 1976) 999 UNTS 171 (ICCPR).
Finally, the prohibition against forced labour oriented in the 1930 Forced Labour Convention
– to which 177 States are party, and which demands the criminalisation of illegal exaction of forced or
compulsory labour at article 25 – has achieved more effective legislative uptake.
Although 44 States
appear not to have enacted the prohibition, 104 have constitutional provisions, 81 have penal
provisions and 58 have labour provisions (for a total of 149 States with some form of legislation).
However, the punishments attaching to these provisions are very limited in a large number of States,
with many involving the imposition of relatively small fines and no further consequences regardless
of the circumstances of the case. Such measures are unlikely to qualify as ‘really adequate’ penalties
as required by article 25 of the Forced Labour Convention, and it appears that the obligation of strict
enforcement suffers from a meaningful implementation gap.
Aside from simple implementation figures, this research (and therefore the dataset) allows
for comparative analysis of domestic implementation across a range of variables which may provide
insight into the correlation between identified factors and domestic implementation measures. For
the purpose of brevity, regional arrangements will be used as a demonstration of this capability which
might also address a wide range of other factors. In considering the geographic distribution of
domestic provisions prohibiting the four identified forms of human exploitation, it is surprising to note
that there are no significant trends which differentiate the five identified regional groups (see fig. 3).
All regions perform significantly better in the implementation of the prohibition against human
trafficking – unsurprising given its near universal enactment. The remaining analysis will therefore
focus on the other three identified forms of exploitation, in which different regions fluctuate in their
standing in a hierarchal ranking and no region comes out as superior across the board.
Above n 8.
In this analysis, regional groups are based on United Nations Regional Groups, a geopolitical rather than purely
Latin America and the Caribbean, and Western Europe and Others evidence a preference for
a particular form of prohibition, with the former favouring constitutional provisions and the latter
criminalisation across all three forms of exploitation. Asia-Pacific and Eastern Europe favour
criminalisation in relation to slavery and servitude (although only by a relatively small margin for
servitude) but reverse this preference in forced labour where constitutional provisions are favoured.
Although the fluctuations in the African regional group are less stark, there is a minor preference for
constitutional provisions in relation to slavery and servitude, which is switched in the context of forced
labour which finds its place in criminal law more frequently. This lack of consistent domestic
implementation across different forms of exploitation is mirrored in the consideration of parties to
regional courts – often identified as a significant influence in the implementation of human rights
standards (see fig. 4). Thus, although membership in such judicial bodies may significantly impact
other implementation measures beyond legislative prohibition, as well as the effectiveness of
domestic legislation, they do not significantly impact the fulfilment of the basic obligations to prohibit
and criminalise human exploitation.
Africa 0.57 0.48 0.5 0.35 0.52 0.61 0.52 0.94
Asia-Pacific 0.39 0.67 0.19 0.24 0.61 0.46 0.35 0.98
Eastern Europe 0.13 0.57 0.17 0.35 0.87 0.22 0.26 1
Latin America & Caribbean 0.7 0.3 0.61 0.24 0.49 0.27 0.12 0.97
Western Europe & Others 0.1 0.59 0.07 0.41 0.24 0.31 0.03 0.93
Fig 3. Domestic Legislation by Region
Legal prohibition and criminalisation constitute only the first step in a broader process of
bringing extreme forms of human exploitation to an end globally and meeting international
obligations in this respect. The presence of legislation alone cannot serve as a reliable indicator of
effective implementation of international obligations, for a number of States with well-drafted
statutes sit at the top of indicators measuring the prevalence of ‘modern slavery’.
In many ways,
effective implementation of laws is more important than their text, ensuring that the articulated goals
are enacted in reality and not merely on paper. Yet it is upon these texts that abolition relies. Without
legislation, the coercive power of the State cannot be turned against slavery. Without proper
definitions, forms of exploitation can slip through the cracks and persist unchallenged in societies that
have grown accustomed to them. Without criminalisation, exploiters cannot be forced to bear the
consequences of their actions. Without a legal framework, victims have no avenue for redress or
justice. Mapping domestic prohibition, and the implementation gap in this field, is therefore vital to
the movement against extreme forms of human exploitation and may also speak to the role and
influence of international legal norms.
India, for instance, has human trafficking provisions which reflect the Palermo definition, comprehensive
bonded labour legislation, and (historically) criminal offences concerning slavery, yet sits at the top of global
prevalence measures for ‘modern slavery’ with an estimated 1.4% of the population ‘in modern slavery’. See the
Walkfree Foundation, ‘Findings’ (Global Slavery Index 2016) <www.globalslaveryindex.org/findings>.
Slavery Servitude Forced Labour Human Trafficking
Fig 4. Domestic Implementation by Regional Court Parties
ACHPR ECHR IACHR All States (control)