© The Author (2008). Published by Oxford University Press. All rights reserved.

For permissions, please e-mail: journals.permissions@oxfordjournals.org.

doi:10.1093/ijrl/eem068, Advance Access Published on February 16, 2008

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

Redundant or Essential? How Politics
Shaped the Outcome of the 1967 Protocol

SARA E. DAVIES*

Abstract
The 1967 Protocol Relating to the Status of Refugees has been described as an unnecessary
addendum to the 1951 Convention Relating to the Status of Refugees. However, if the
1967 Protocol was superfluous, why did the United Nations High Commissioner for
Refugees in the early 1960s insist on its development? This article seeks to establish that the
1967 Protocol was originally intended to encompass the broader concerns of African and
Asian states concerning refugee populations in their region. However, the political influence
upon the development of international refugee law radically altered the UNHCR’s
endeavour to make the 1951 Convention universally accessible.

1. Introduction
The 1967 Protocol Relating to the Status of Refugees (hereafter referred
to as the 1967 Protocol) was negotiated in only three years, but the need
for a supplement to the 1951 Convention Relating to the Status of
Refugees (hereafter referred to as the 1951 Convention) was, arguably,
identified in the 1950s. By the mid-1960s, the Office of the United
Nations High Commissioner for Refugees (UNHCR) had grown
increasingly frustrated with its limited mandate, with growing refugee
problems in Europe and Africa that were beyond the technical reach of
the Convention, and with a lower than expected number of accessions to
the Convention, particularly amongst recently decolonized states. For the
UNHCR, the 1967 Protocol was seen as a tool that might remedy these
problems.1 It was formulated as an Addendum to the 1951 Convention,
yet was a separate legal instrument developed with the intention of permitting
states to adopt the 1967 Protocol and still be bound by the 1951
Convention (eliminating the cumbersome need of separate accession to
the latter). The key difference between the 1951 Convention and 1967
Protocol was the removal of the 1951 Convention’s time and geographic
constraints. Within Article 1A of the 1951 Convention, a refugee was
defined as an individual with a ‘well-founded fear of persecution’ due to
events in Europe prior to 1 January 1951, with the option for states to

* Dr Sara E. Davies is a lecturer at the School of Justice, Faculty of Law, Queensland University of
Technology, Australia. Email correspondence is welcome at: s5.davies@qut.edu.au.
1 UNHCR, Explanatory Note, handed to Mr Goormaghtigh of the Carnegie Endowment on 24 Sept.
1964, Folio 24, 16/1/3 AMEND, Series 1, Fonds 11.
704 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

keep the geographic constraints or to remove it from their accession
document.2
Because the 1951 Convention had the option for states to remove or
keep the geographic constraints, there has been a long-standing insistence
that the 1967 Protocol was unnecessary. This argument was posed, most
recently, by Kazimierz Bem.3 Bem argued that because the Convention
had an option for acceding states to remove geographic constraints and the
discrepancy between the 1951 Convention and UNHCR Statute (to be
discussed in greater detail later) was not raised until 1965, there were few
grounds for High Commissioner Schnyder’s insistence on a new Protocol.
According to Bem and Jackson, amongst others, the 1967 Protocol was
unnecessary precisely because states were permitted to accede to the 1951
Convention but derogate from the geographic constraints in Article 1B.4
Furthermore, they argue that European states were, in practice, more than
willing to act to protect refugees regardless of the time and geographic
constraints.5 In a similar vein, Goodwin-Gill insisted that the 1967 Protocol
was legally unnecessary because, even if there was concern about the initial
restriction of the 1951 Convention through these time and geographic
constraints, it was ‘superficial’ to claim that the Convention was Eurocentric
because all UN member states were invited to send representatives to the
Conference of Plenipotentiaries charged with drafting the Convention.6
This article puts forward an alternative understanding of the 1967
Protocol. It argues that High Commissioner Schnyder intended a different
Protocol to the one created, one that would be more acceptable to a wider
number of states and applicable to a wider range of refugee causes than the
rather narrow, Eurocentric definition contained in Article 1A and 1B.7 First,
the UNHCR wanted to create the 1967 Protocol in order to demonstrate
that the time and geographic constraints within the 1951 Convention would
no longer affect the operations that the UNHCR could undertake outside
Europe. As we will see, this became particularly important in light of the
contrasting response by the UNHCR in the 1950s to the Hungarian refugee
population and to the Chinese and Algerian crises. Second, by the 1960s the

2 Article 1B specifies that the application of the definition of the ‘refugee’ in Article 1A of the 1951
Convention is either (a) ‘events occurring in Europe before 1 January 1951’; or (b) ‘events occurring in
Europe or elsewhere before 1 January 1951’ (emphasis added). P. Weis, The Refugee Convention 1951: The
Travaux Preparatoires Analysed, with a Commentary by the late Dr Paul Weis (Cambridge: University of
Cambridge, 1995), xiii-xiv.
3 K. Bem, ‘The Coming of a “Blank Cheque” – Europe, the 1951 Convention, and the 1967
Protocol’, 16 IJRL 609 (2004), 609-627.
4 I. C. Jackson, The Refugee In Group Situations (The Hague: Martinus Nijhoff, 1999); K. Bem, ibid.
5 K. Bem, above n. 3.
6 G. Goodwin-Gill, The Refugee in International Law (Oxford: Clarendon Press, 1996), 297, n.3.
7 UNHCR, High Commissioner Schnyder to Mr Stavropoulos, Memorandum by the United Nations
High Commissioner for Refugees on the Report of the Colloquium on Legal Aspects of Refugee Problems held in April
1965 in Bellagio (Como), Italy, 6 Aug. 1965, Folio 136, 16/1/3 AMEND, Series 1, Fonds 11.
 The Shaping of the 1967 Protocol 705

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

UNHCR faced mounting pressure from newly decolonized African and
Asian states to reform the 1951 Convention, which was widely perceived to
be a Eurocentric instrument. In order for the UNHCR to build a universal
profile across all regions the 1967 Protocol was a necessary creation to dispel
the growing concerns by developing states about the applicability of the
1951 Convention to their refugee situations.
The argument presented in this article is that the UNHCR, in particular
High Commissioner Schnyder, failed in the pursuit of a expansive Protocol
largely because of what Rajagopol terms ‘radically pluralist interpretations’
of the law.8 What we saw in the case of the 1967 Protocol, was that Schnyder’s
initial concept changed radically as political objectives merged with the legal
discussion regarding the 1951 Convention and its new Protocol. Scholars
such as Bem, Jackson and Goodwin-Gill are correct to point out the potential
for expanding the scope of the 1951 Convention, without perhaps a need for
the 1967 Protocol. However, High Commissioner Schnyder argued that the
UNHCR did not only need to overcome its assistance problems, but that
there were political and legal circumstances surrounding the 1951 Convention
that also needed to be changed. In particular, there was a need to overcome
the dominant interpretation of the 1951 Convention by newly decolonized
states, who argued that the Convention only had European application,
through the creation of the 1967 Protocol. International refugee law was
immersed in the process of what critical legal scholars describe as ‘legal
argumentation’,9 whereby the struggle between politics and law leads to
exclusion, in order to privilege particular concepts that are more politically
acceptable to powerful actors. As this article will demonstrate, we see the
UNHCR engaged in legal argumentation with the outcome being that this
organization was unable to persuade the more powerful states to see the
merit of creating a new, more widely encompassing Protocol.
This article contends that the UNHCR attempted to articulate a new
interpretation of international refugee law in the 1960s. Its primary focus
on post- Second World War and Cold War refugee crises in Europe led to
a belief, amongst decolonized states, that the UNHCR’s utility outside of
Europe was very limited. For an institution which needed universal
applicability in order to survive and thrive, particularly in light of the
growing need for refugee assistance in these new, troubled regions, this
perception had the potential to serve as a death knell for the UNHCR.10
Therefore, the need to constitute a new role for the UNHCR was considered
8 B. Rajagopol, International Law from Below: Development, Social Movements and Third World Resistance
(Cambridge: Cambridge University Press, 2003).
9 D. Kennedy, ‘A New Stream of International Law Scholarship’ (1988–1989) 7 Wisconsin International
Law Journal 1-50; M. Koskenniemi, From Apology to Utopia: The Structure of International Legal Argument
(Helsinki: Finnish Lawyers’ Publishing Company, 1989), xvi, 40-51.
10 For an in-depth discussion of the struggles that UNHCR faced in its development, see,
G. Loescher, The UNHCR and World Politics: A Perilous Path (Oxford: Oxford University Press, 2001), Chs.
3-4.
706 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

paramount – the 1967 Protocol was to be a part of framing this new role,
through the creation of a law that would open new political possibilities.
These arguments depend on how we understand international law.
According to Koskenniemi, ‘in law, political struggle is waged on what
legal words … mean, whose policy they will include, and whose they will
exclude’.11 In light of this understanding, the notion that the 1967 Protocol
was unnecessary carries weight, if we accept that removing the time and
geographic constraints was the only purpose behind the Protocol. However,
this article will argue that this was not the primary objective of the
UNHCR. Therefore, it is crucial that we look at how the political struggle
that waged over the creation of the 1967 Protocol developed in the way
that it did; and how this shaped both the Protocol itself and its interpretation.
In tracing the events that shaped the creation of the 1967 Protocol, and
that caused its final form to move further and further away from that
intended at its initial creation, we need to follow the arguments of those
who wanted the 1967 Protocol in the first place, in order to understand
why they advocated such an approach.
The remainder of this article develops three main points. First, it will look
at the difficulty that the UNHCR was having reconciling the 1951 Convention’s
refugee definition with the broader definition of a refugee contained in its
mandate, and how the growing number of ‘mandate refugees’ as opposed to
‘Convention refugees’ prompted the UNHCR to support changes to the
Convention. Second, in Europe, the 1960s saw an end to the legacy of
refugees from the Second World War but a continuation of refugee flows
from those fleeing communism in Eastern Europe. At the same time, the
number of refugees in Africa was growing rapidly as a result of violent
struggles for decolonization and its accompanying civil wars.12 The Asian
region was also experiencing a higher number of refugees caused, in particular,
by the emergence of a communist regime in China and the breakup of
French Indochina. Finally, this led to the problem of assistance. While the
UNHCR was able to provide material assistance, it was not mandated to
provide legal or protective assistance because the region’s states had not
acceded to the Convention and the region’s refugees did not meet the
Convention’s definitional requirements.13 In contrast to the refugees seeking
refuge in Europe, refugees determined to fall outside the Convention’s refugee
definition were assisted as ‘mandate refugees’ under the ‘good offices’ practice
of the UNHCR. The problem the UNHCR faced with ‘mandate refugees’
was that legal protection and durable solutions, such as resettlement or

11 M. Koskenniemi, ‘International Law in Europe: Between Tradition and Renewal’ (2005) 16 The
European Journal of International Law 119.
12 Jackson, above n. 4, 143-175.
13 L. Holborn, Refugees: A Problem of Our Time (New Jersey: Scarecrow Press, 1975), 657-659;
Loescher, above n. 10, 188.
 The Shaping of the 1967 Protocol 707

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

repatriation, were difficult to achieve because this kind of assistance required
cooperation with other states, and that meant extending the UNHCR’s efforts
and the responsibilities of states towards refugees beyond the scope of the
Convention. The interconnection of these three developments prompted
High Commissioner Schnyder to put forward the case for a new Protocol
and, as the remainder of the article demonstrates, the politico-legal struggle
that ensued led to the 1967 Protocol reflecting little of its original intention.

2. UNHCR and the concept of ‘old’ and ‘new’ refugees

According to Bem, the Protocol was created by ‘generous’ European
states acting for humanitarian reasons in order to extend ‘international
legal protection to other refugees’,14 having sorted out the European
refugee problems caused by the Second World War and the spread of
communism. However, this argument is somewhat contradictory as Bem,
as well as Jackson and Goodwin-Gill, argued that the 1967 Protocol was
an unnecessary piece of legislation, insisting that the Convention already
afforded protection to all refugees being assisted by the UNHCR.15 James
Hathaway, among others, endorses the idea that the 1967 Protocol did
finally universalise the 1951 Convention, but he insists that the 1967
Protocol did not go far enough in changing the substance of the Eurocentric
instrument.16 Such authors point out that the character of the 1951
Convention’s refugee definition and the treatment of refugees by host
states was, and still is, submerged in Eurocentric concepts and practices.
Hathaway’s point will be taken further by assessing whether a privileging
of European concepts can be traced and, through archival material, a
distinct history of the legal argumentation process behind the 1967
Protocol will be presented. Although Hathaway alludes to the 1967
Protocol remaining Eurocentric, through exhaustive empirical research
this article reveals why the UNHCR lobbied for the 1967 Protocol and
outlines the difficulties it encountered in its attempt to make the Protocol
substantially different from the 1951 Convention.
The UNHCR’s first aim in lobbying for a Protocol was the universalisation
of international refugee law and, in turn, the universalisation of its own
mandate. The Protocol removed the temporal and geographic constraints
that had prevented the UNHCR from taking a greater role in refugee

14 Bem, above n. 3, 627.

15 Bem, above n. 3; Jackson, above n. 4; Goodwin-Gill, above n. 6.
16 E. Lentini, ‘The Definition of Refugee in International Law’ (1985) 5 Boston College Third World
Law Journal; A. Zolberg, A. Suhrke and S. Aguayo, Escape From Violence: Conflict and the Refugee Crisis in the
Developing World (New York: Oxford University Press, 1989), 25-31; J. Hathaway, The Law of Refugee
Status (Toronto: Butterworths, 1991), 10-11; A. Grahl-Madsen, ‘Refugees and Displaced Persons:
Meeting the Challenge’, in P. Macalister-Smith and G. Alfredsson (eds.), The Land Beyond: Collected
Essays on Refugee Law and Policy by Atle Grahl Madsen (The Hague: Martinus Nijhoff Publishers, 2001),
318.
708 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

protection. But the Protocol did not address what High Commissioner
Schnyder saw as the UNHCR’s greatest difficulty at that time, namely,
how to provide protection for groups of refugees, comprising individuals
who could not prove that they faced individual persecution, and those who
fled due to more generalised violence/persecution.17 Schnyder perceived a
need for a new instrument that could provide large groups of asylum
seekers with prima facie determination.18 However, the Protocol did not
address the tension between refugee protection for those that fled en masse,
often amidst armed conflicts (the norm for refugee movements in the
developing world),19 and protection for those that had individual reasons
for flight (the norm that was emphasised in the 1951 Convention).20 Nor
did the Protocol directly address the fact that the 1951 Convention
contained bureaucratised standards and procedures for host states in
relation to the determination and treatment of refugees. Arguably, this
problem was exacerbated by the way states chose to implement the 1951
Convention, which does not necessarily insist upon such a bureaucratised
approach to refugee status determination. In developing states, such
capacity is often lacking. Moreover, refugees often come under the domain
of the state’s most organised bureaucratic structure, the military, which
tend to see refugees as a security, rather than humanitarian, problem.21
The outcome was that the 1967 Protocol overlooked an inherent tension
that the UNHCR had been attempting to deal with since the 1950s. This
tension, of how to deal with the type of refugee influx commonly
confronting states in Africa and Asia, as opposed to the types of influx that
confronted developed states in Europe and North America, created an

17 UNHCR, Memorandum, above n. 7; UNGA, Official Records of the Third Committee, Twenty-First
Session, 1447th Meeting, 5 Dec. 1966, 416.
18 The High Commissioner wrote in 1966 that he had ‘given further consideration to a further legal
problem, which arises in connexion with the substantive definition of the term “refugee” in the 1951
Convention. Experience, and more especially recent experience of new refugee situations in Africa,
has shown that certain States, for various reasons, may not be able to resort to individual eligibility
determination … Having regard to this experience, it might be of advantage if the existing definition
was supplemented by certain criteria for the prima facie determination of such group eligibility’.
UNHCR, Memorandum, above n. 7, 6-7. The need for individual status determination still exists today
for those in fear of refoulement, as demonstrated by Michael Alexander and Michael Kagan.
M. Alexander, ‘Refugee Status Determination Conducted by UNHCR’, 11 IJRL 251-289 (1999);
M. Kagan, ‘The Beleaguered Gatekeeper: Protection Challenges Posed by UNHCR Refugee Status
Determination’, 18 IJRL 1-29 (2006).
19 Zolberg, et al., above n. 16; Hathaway, above n. 16; S. Schmeidl, ‘Conflict and Forced Migration:
A Quantitative Review’, in A. R. Zolberg and P. M. Benda (eds.), Global Migrants, Global Refugees: Problems
and Solutions’, 2001.
20 M. McGuiness, ‘Legal and Normative Dimensions of the Manipulation of Refugees’, in
S. J. Stedman and F. Tanner (eds.), Refugee Manipulation: War, Politics, and the Abuse of Human Suffering,
2003.
21 For more on this point, see, G. Chatelard, ‘Jordan as a transit country: semi-protectionist
immigration policies and their effects on Iraqi forced migrants’, New Issues in Refugee Research, Working
Paper No. 61, (Geneva: UNHCR Research Unit, 2002).
 The Shaping of the 1967 Protocol 709

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

important division. The division was between those states that wanted the
UNHCR to develop a role in protecting mass influx refugees, and those
that insisted the onus remain on the individual to prove that they were
targeted for persecution. The direction of the Protocol itself was the result
of a choice made by the UNHCR to avoid the creation of a larger, more
nuanced, Convention, which might have tackled this fundamental question,
and to instead lobby for a Protocol that simply removed the Convention’s
time and geographic constraints, which were preventing its universal
application. Therefore, although the question of mass influx response was
certainly raised by the UNHCR prior to the drafting of the Protocol, in
the end the Protocol did not address the issue.
As the UNHCR grew in organisational strength, the potential for it to
provide more widespread protection also increased. By 1954, the UNHCR
was emerging from a financial crisis that almost saw it close down. The
organisation was saved by a Ford Foundation grant in 1952 of approximately
$2.9 million.22 In the early 1950s, High Commissioner Gerrit Jan van
Heuven Goedhart presented a number of speeches to the General Assembly
insisting that protection without material assistance – the image of the
UNHCR set out in the Convention – was an incomplete mechanism for
addressing the problem of refugees. Goedhart pointed out that there were
400,000 refugees who remained in camps throughout Europe because they
had been rejected for resettlement on account of age, illness or disability.23
He stressed the need to integrate these refugees and, on 2 February 1952,
the General Assembly agreed to allow Goedhart to raise funds through
member state donations to provide the UNHCR with the capacity to provide
material assistance to these refugees. The General Assembly permitted
Goedhart to raise $3 million for this purpose but, by March 1954, he had
secured only $1 million in donations. Meanwhile the US had donated $45
million to the Intergovernmental Committee for European Migrants.24
When the UNHCR obtained the Ford Foundation grant in 1952, the
Office was able to involve itself, for the first time, in providing administrative
assistance to refugees in Western Europe and was able to play a role in

22 Donor states opposed the UNHCR being able to raise its own funds for a variety of operational
purposes. European states wanted to put all of their funds, including foreign aid, into the local
integration of their refugees. France, in particular, made every effort to create problems for the
UNHCR as their French UN Representative, Robert Rochefort, had aspired to be the Deputy High
Commissioner, with a US representative as High Commissioner. Meanwhile, the United States had set
up the Intergovernmental Committee for European Migration (ICEM, now the International
Organisation for Migration [IOM]) which managed the migration of refugees and labour migrants;
was fully funded and operated by the United States outside of the UN; and only democratic nations
friendly to the US were on the Board. ICEM was directly concerned with those fleeing communist East
Europe. As a result, the US would not fund the UNHCR. See, Holborn, above n. 13, 387; Loescher,
Beyond Charity: International Cooperation and the Global Refugee Crisis (Oxford University Press, 1993), 62-63;
Loescher, above n. 10, 61-62, 67.
23 Loescher, above n. 10, 63.
24 Loescher, ibid., 64.
710 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

coordinating the activities of NGOs in this region.25 The Ford donation
encouraged other donations – including donations from governments –
marking an important shift in the UNHCR’s operations from a
preoccupation with legal protection towards a greater interest in assistance.26
The UNHCR’s widely perceived success in directing material assistance to
refugee groups during the Berlin crisis in 1953 helped to further this shift.27
The Berlin experience encouraged the General Assembly to approve an
expanded financial budget for the UNHCR in 1954, which included a
significant allocation for assistance. Member states were to be part of a
United Nations Refugee Emergency Fund (UNREF) Executive Committee28
(now the UNHCR Executive Committee). The UNREF Executive
Committee would then, as it does now, advise the High Commissioner on
protection and assistance programmes and play a ‘direct over-sight role
over the Office’s finances’.29 The Executive Committee could directly
control the UNHCR’s activities and, as a result, the UNHCR began to
secure a much higher proportion of its funding from member states as
opposed to private institutions.
The improving financial situation and increased state support prompted
the UNHCR to attempt to expand its mandate. China, then represented
by the Nationalists, requested assistance for those fleeing communist China
as early as 1952. In Europe, Austria and Yugoslavia asked for UNHCR
assistance with Hungarian refugees in 1956, following their foiled revolt
against communism. Tunisia and Morocco sought assistance from the
UNHCR with a large number of Algerians fleeing the civil war. However,
the limits of both the UNHCR’s mandate and the Convention were
revealed by the variable treatment these cases received (see below).
A ‘Convention’ refugee was one recognised as fleeing events in Europe
prior to 1 January 1951 due to an individual fear of political persecution.
A ‘mandate’ refugee was one generally recognised to need the assistance of
the High Commissioner’s ‘good offices’ because they were fleeing political
persecution. But a ‘mandate’ refugee did not fall under the time and
geographic conditions in the 1951 Convention (or, in most cases, as
individuals who could prove their direct fear of political persecution).30

25 Soguk, States and Strangers: Refugees and Displacements of Statecraft, 1999, 172.
26 Holborn, above n. 13, 324.
27 In 1953, the number of East Germans fleeing went from 15,000 in Dec. 1952 to 48,000 in Mar.
1953, due to fear of increasing Sovietisation of East Germany. Goedhart used his Ford funding to assist
with temporary housing, assistance and voluntary agencies to deliver assistance. J. Read, The United
Nations and Refugees – Changing Concepts (Geneva: Carnegie Endowment for Peace, 1962).
28 General Assembly Resolution 5388 (VI) and 832 (IX) respectively.
29 Loescher, above n. 10, 70.
30 For more on this, see, UNHCR, Memorandum by the United Nations High Commissioner for Refugees on the
Report of the Colloquium on Legal Aspects of Refugee Problems held in April 1965 in Bellagio (Como), Italy, 23 Sept. 1965,
HCR/RS/31, Series 1, Fonds 11, 16/1/3 GEN, 5; Goodwin-Gill, above n. 6; P. Kourula, Broadening the
Edges: Refugee Definition and International Protection Revisited (The Hague: Martinus Nijhoff Publishers, 1997).
 The Shaping of the 1967 Protocol 711

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

Therefore, these groups received different levels of assistance. It is significant
that, in the late 1950s, out of the three major groups to seek UNHCR
refugee assistance through the Executive Committee (excluding the post-
WWII population), the Hungarians were given ‘Convention’ refugee status,
while the two classified as ‘mandate’ refugees were outside Europe (Chinese
and Algerian refugees). This outcome prompted the third High
Commissioner, Felix Schnyder, to stress, in 1961, the need to abandon the
notions ‘old’ (traditional European/‘Convention’ refugees) and ‘new’
(other/ ‘mandate’ refugees) refugees. Schnyder argued that both groups
should be treated similarly by the UNHCR and offered access to equal
levels of legal protection and material assistance.31 It is at this stage that we
see the initial process of legal argumentation developing, due to the tension
that emerged between the political context in which the 1951 Convention
initially excluded some groups from refugee status, and the new situations
that required the instrument to be less exclusionary. These three cases will
now be briefly considered in turn.

2.1 China
This prompts the question of how the separation between ‘Convention’
and ‘mandate’ refugees was understood in the 1950s and 1960s, and,
most importantly, how this distinction impacted upon the formation of
the 1967 Protocol. At the UN General Assembly in 1952, China requested
that the hundreds of thousands of Chinese fleeing communism since
1945 be considered refugees and that their plight be brought to the
attention of the UNHCR.33 From the High Commissioner’s perspective,
however, this request was problematic.34 There were several reasons for
this. First, the existence of two Chinas meant that Nationalist China,
which was the then internationally recognised China, presented a national
territory in which people were safe to reside and, thus, the refugees could
not be identified as such under the 1951 Convention.35 However, the
sheer scale and cost of the Chinese refugee problem meant that they
could not all live in Taiwan, thus further complicating the situation if this
was regarded as the sole solution.36
The second problem related to British administered Hong Kong, where
most of the Chinese had sought asylum. The UK opposed the provision of

31 Jackson, above n. 4, 102-106.

32 UNHCR, Memorandum, above n. 30; Kennedy, above n. 9; Koskenniemi, above n. 9.
33 UNGA, Colloquium on the Legal Aspects of Refugee Problems, Note by the High Commissioner,
Executive Committee of the High Commissioner’s Programme, Thirteenth Session, 5 May 1965,
A/AC.96/INFO.40, Annex 1, 1. Rockefeller Foundation Archives, Record Group 1.2, Series 200, Box
484, Folder 4143.
34 Holborn, above n. 13, 657-658, 676-677.
35 Jackson, above n. 4, 92.
36 Loescher, above n. 10, 93.
712 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

UNHCR assistance for Chinese refugees, primarily because it believed
that such assistance would encourage further flight. Britain had not signed
Hong Kong on to the 1951 Convention and, thus, the territory had no
obligation to the Convention or to any refugee arrivals. Britain argued that
the People’s Republic of China (PRC) should be recognised as a legitimate
state and, therefore, Chinese refugees did have a legitimate state of origin
to return to, and that doing so would not pose a threat to their personal
security.37 A third problem was caused by the fact that the Chinese refugees
had fled en masse and the UNHCR lacked both the funds and the
administrative capability to screen each individual. Moreover, it was far
from apparent what would happen to these individuals even if they were
screened and found to be refugees, with Taiwan unable to accept them and
Britain not willing to.38
Within this context, High Commissioner Goedhart wanted to determine
whether the UNHCR could potentially assist the Chinese as refugees and
he commissioned a study on the Chinese who had managed to enter Hong
Kong and the nature of their situation. The study was duly approved by
the Advisory Committee of the High Commissioner’s Programme (now
the Executive Committee).39 It found that ‘it seemed safe to assume that
fear of persecution in a country in the throes of a revolution was seldom
without foundation’ and, therefore, ‘objectively speaking it could be
accepted that a well-founded fear of persecution was not excluded’.40
Despite these findings, the problem of the ‘two Chinas’, and the British
refusal to accept that the UNHCR’s mandate covered the Chinese refugees
in Hong Kong, meant that the refugees did not fall under the High
Commissioner’s mandate and were not eligible for assistance. When the
so-called Hambro report was released in 1955, the Executive Committee
placed ‘considerable emphasis on the need to address the humanitarian
problems of this group without regard to purely legal considerations’.41 In
1957, the new High Commissioner, Auguste Lindt, placed the issue of the
Hong Kong refugees on the Executive Committee’s agenda again; it was
decided that although this population did not fall under the mandate of
the UNHCR the problem was of ‘concern to the international
community’.42 As a result, the High Commissioner was authorized by the
General Assembly to ‘use his good offices to encourage arrangements for
contributions’.43 The Chinese refugees were, therefore, included in this
37 Loescher, ibid., 93-94; Holborn, above n. 13, 376. PRC was formally recognised as the legitimate
state of China by the UN in 1974.
38 Jackson, above n. 4, 91.
39 The study was carried out by Dr Edvard Hambro, The Problem of Chinese Refugees in Hong Kong
(Leiden: A W Sijthoff, 1955). Jackson, above n. 4, 90-91; Loescher, above n. 10, 93.
40 Jackson, above n. 4, 91.
41 Ibid., 92.
42 Ibid., 93.
43 Ibid.
 The Shaping of the 1967 Protocol 713

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

new UNHCR fundraising arrangement. However, it should be noted, in
passing, that the UNHCR did not have much success in raising contributions
to assist the Chinese refugees. For instance, there is evidence to suggest
that, even had these refugees fallen under the remit of the 1951 Convention,
Western states would not have provided resettlement places for them. What
this case illustrates is that the UNHCR Statute could assist refugees who,
for one reason or another, fell outside the Convention’s definition of a
refugee. However, the British refusal to allow the UNHCR to address this
situation fully incapacitated the UNHCR’s work. In addition, despite the
findings of the Hambro report documenting that fear of political
persecution was the reason for people fleeing mainland China, the fact
that flight had taken place in such large numbers meant that the political
and practical repercussions overrode humanitarian concerns.

2.2 Hungary
The failed Hungarian revolt, followed by the Soviet invasion and
occupation in November 1956, led to the exodus of some 200,000
Hungarians to Austria and Yugoslavia.44 These were relatively well-
educated and/or prosperous refugees fleeing communist oppression and
there was little opposition amongst Western states to these refugees
receiving UNHCR assistance and legal protection.45 Unsurprisingly then,
refugee advocates had little difficulty persuading member states that these
refugees fell under the UNHCR Statute. This refugee population was
also immediately granted prima facie refugee recognition under the 1951
Convention. After individual screening in a state that had acceded to the
1951 Convention, all were eligible for resettlement.46 European states,
plus the US, Australia, Canada, Israel and Latin American states, agreed
with the UNHCR’s conclusion and provided resettlement places for the
Hungarians.47 That the Hungarians could be declared ‘Convention
refugees’, despite the revolt taking place in 1956, was a product of legal
gymnastics. It was argued that the Hungarian exodus was indeed a result
of ‘events prior to 1951’ because the Communist Party had established a
People’s Republic in Hungary in 1947–48.48
According to Loescher, the Hungarian situation gave rise to a realisation
that the UNHCR had a similar duty to grant prima facie recognition, where
applicable to the Statute, in other situations of political upheaval –
regardless of where and when this occurred.49 In practice, however, there
remained a sharp distinction between the treatment of refugees in Europe

44 Holborn, above n. 13, 391.

45 Loescher, above n. 10, 87.
46 Jackson, above n. 4, 119.
47 Loescher, above n. 10, 86.
48 Holborn, above n. 13, 394; Loescher, above n. 10, 86.
49 Loescher, ibid., 98.
714 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

and those in the rest of the world, and between those events which could
be construed as having occurred before 1 January 1951 and later events.
This is evident in the different international responses to the Chinese
refugees and the Hungarian refugees. However, during debates about how
best to respond to Algerian refugees, the UNHCR made significant
headway towards balancing this tension.

2.3 Algeria
A war was fought between the Algerian nationalist movement and colonial
France from 1954 to 1962. In 1957 it reached new levels of brutality,
prompting the exodus of thousands of refugees into neighboring Tunisia
and Morocco. Both of these states requested UNHCR assistance,
becoming the first developing states to do so (in itself a fact worth noting).
Both countries had recently become independent of France and were
supporters of the Algerian independence movement.50 In sharp contrast
with the response to the Hungarian refugee crisis, the UNHCR was more
circumspect in the Algerian case. It ensured that, although the Algerians
were referred to as ‘refugees’ in internal discussions, and although in his
private communications with governments the High Commissioner
confirmed that the refugees were prima facie refugees according to the
1951 Convention, there were no public statements confirming this.51 As
a result, the Algerians were not publicly recognised as ‘Convention’
refugees. Clearly, this was partly a result of the political tensions
surrounding the Algerian War. The movement of refugees had also
occurred en masse, making individual status determination practically
impossible. Despite this, by the end of 1958, the UNHCR was officially
providing material assistance to help Tunisia and Morocco with the
Algerian refugee population.52
The conduct of the war by the French was notoriously brutal. In addition
to ransacking villages, the French forcibly resettled peasants in guarded
encampments and made systematic use of torture. This brutality persuaded
High Commissioner Lindt to argue that there was a strong humanitarian
imperative to respond to Tunisia and Morocco’s requests for help.53
However, the level of assistance that Lindt could offer was tempered by the
political context. In short, it was politically unfeasible to declare that
France, a permanent member of the UN Security Council, was conducting
50 Loescher, ibid., 97-99.
51 See Jackson, above n. 4, 120-142, for a more in-depth appraisal of the Algerian refugee status
debate. Also, C. Ruthstrom-Ruin, Beyond Europe: The Globalization of Refugee Aid (Lund: Lund University
Press, 1993).
52 The resolution put forward by Tunisia and Morocco for UNHCR assistance due to the urgency
of the Algerian refugee problem was, interestingly, supported by Libya, Pakistan, Sudan, Cambodia,
Turkey, Japan, United Arab Republic, USSR, Albania, Liberia, Ethiopia, Haiti and Indonesia. Jackson,
above n. 4, 126-127.
53 Loescher, above n. 10, 97-98.
 The Shaping of the 1967 Protocol 715

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

political persecution in Algeria to the extent that those fleeing could be
considered prima facie refugees.
Lindt continued to insist, though, that the UNHCR had to respond to
the requests for assistance by Tunisia after the first arrival of 85,000
Algerians. Lindt questioned how the UNHCR could provide prima facie
recognition to Hungarians and not to Algerians, simply because of the
wider political context. Furthermore, he feared that, if the UNHCR did
not respond, the Office would forever be seen by developing countries as
‘the High Commissioner for European refugees only’.54 Importantly, the
US, by now a major donor to the UNHCR, supported UNHCR assistance
to the Algerian refugees, as long as it was done discretely. The US calculated
that assistance delivered through the UNHCR would represent a show of
good will to African states and would help prevent the spread of communism
in North Africa. As a result, and despite the potential consequences, the
UNHCR decided to provide prima facie refugee status to the Algerian
refugees under the ‘good offices’ section of the UNHCR Statute. This was
an important precedent and highlighted the UNHCR’s sensitivity to
developing states’ perception of it as a ‘European institution’ and to the
need to change that image.55 However, the Algerian case also demonstrates
some of the powerful constraints that were placed on the UNHCR in this
endeavor.
In the early to mid 1960s, there was a rising number of refugees in
developing states, in particular, Africa and Asia. This was largely due to
wars caused by armed struggles for decolonisation, border disputes, and
civil conflicts. As a result, refugee problems in Europe started to pale in
comparison with the mass influx of refugees occurring in developing
regions. Loescher credits Lindt’s successor as High Commissioner, Felix
Schnyder, who was elected by the General Assembly in 1960, as being the
key player in expanding the UNHCR’s presence outside Europe and in
utilizing the UNHCR’s ‘good offices’ authority to assist refugee groups not
previously addressed by the Statute or Convention.56 The need to constantly
stretch the ‘good offices’ function in this way provided part of the rationale
for eventually abandoning the time and geographic constraints within the
1951 Convention.57

3. UNHCR and refugee crises in the developing world

Another aim of the UNHCR in fostering the Protocol had been to attract
more accessions to the 1951 Convention. This was closely linked to a
third aim: to prevent or forestall the proliferation of regional refugee

54 Loescher, ibid., 98.

55 Jackson, above n. 4, 120-142; Soguk, above n. 25, 173; Loescher, above n. 10, 98-101, 105-110.
56 Loescher, above n. 10, 106.
57 Soguk, above n. 25, 173; Loescher, above n. 10, 106.
716 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

instruments, such as the Organisation of African Unity (OAU) 1969
Refugee Convention, which could take precedence over the UNHCR
and create a system of differential regional refugee protection. Throughout
the 1960s, the UNHCR was called upon to assist hundreds of thousands
of refugees in Africa, under General Assembly Resolution 1673 (XVI) of
1961, which allowed the High Commissioner to extend the mandate
function of ‘good offices’ to ‘groups of refugees’.58 The High
Commissioner’s argument (see above) for the need for the international
community to assist ‘new’ groups of refugees, and not to restrict its
activities to so-called ‘old’ refugees (located in Europe), had finally paid
dividends in terms of who the UNHCR could now assist.
The crises in Africa that the UNHCR became involved in during the
1960s involved Angolan refugees in the Congo; Ghanaian refugees in
Togo; and Rwandan refugees in Tanganyika, Uganda, Burundi and the
Congo. By the mid-1960s, the UNHCR assistance programme in Africa
had expanded to include refugees from Portuguese Guinea in Senegal;
refugees from Mozambique in Tanzania, Zambia, Kenya and Swaziland;
refugees from the Sudan in Uganda, Ethiopia, the Central African Republic
and the Congo; refugees from the Democratic Republic of Congo in
Zambia and the Central African Republic; refugees from Burundi in
Rwanda, in 1966; and refugees from Ethiopia in the Sudan, in 1967. By
the late 1960s there were further movements of refugees from Chad,
Southern Africa and Namibia.59
During this period, the UNHCR was criticised for providing a large
amount of material assistance without concomitant levels of legal
protection. The emphasis on material assistance as the sole solution to the
problems in Africa was widely considered as flawed.60 It certainly served to
strengthen the idea that there was a fundamental difference between
European and non-European refugees; the former were seen to primarily
need legal protection (plus assistance), while the latter were only provided
with material assistance. This characterisation clearly had its roots in the
fact that the UNHCR was initially formed to respond to refugee problems
in Europe alone.61 Even though Europe underwent massive social,

58 Jackson, above n. 4, 104. It needs to be noted that this was assistance only – not protection.
59 For an in-depth appraisal of the legal status and assistance provided to these entire refugee
groups, see, UNHCR, Memorandum, above n. 30; Holborn, above n. 13, 957-1386.
60 High Commissioner Schnyder started emphasising that the ‘new refugees’ needed material
assistance, as opposed to the ‘old refugees’, almost immediately. As a result, legal protection for refugees
in Africa was, obviously, not considered by Schnyder as the solution to the ‘new refugees’. See, UNGA,
Colloquium, above n. 33; Jackson, above n. 4, 103-104.
61 This has been repeatedly argued by Zolberg, et al., above n. 16; J. Hathaway, ‘Reconceiving
Refugee Law as Human Rights Protection’ (1991) 4 JRS 113-131; B. S. Chimni, ‘The Geopolitcs of
Refugee Studies: A View from the South’ (1998) 11 JRS 350-374; Soguk, above n. 25; Loescher, above
n. 10.
 The Shaping of the 1967 Protocol 717

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

economic and border upheavals after both world wars, there was a familiar
sense of laws, administration and pre-existing norms about where people
‘fitted’ and about where those who did not ‘fit’ should go. Europe’s main
refugee concern after the Second World War was to resettle refugees within
a post-war context of relative peace. Africa, especially, confronted an
entirely different situation. There, refugees were fleeing war and being
settled in a context of ongoing violence, often in places where the state was
either not yet fully-formed or also involved in conflict. The 1951 Convention
simply did not provide a realistic mechanism for dealing with refugees in
this context. High Commissioner Schnyder attempted to address this
problem by expanding the ‘good offices’ concept, making it applicable to
the large groups of refugees in Africa, and by focusing on material
assistance; this was all the ‘good offices’ concept could provide. The result,
however, was neglect of legal protection and a failure to address the
insecurity and settlement problems faced by refugees in host countries.
The UNHCR’s practice in Asia prior to the 1967 Protocol was similar
in certain respects to the consideration it gave to African refugee needs.
The emphasis here was also on the notion that the High Commissioner
‘was compelled to create responses different from those developed in his
work in Europe’.62 Therefore, the UNHCR focused on the provision of
basic material assistance and relief measures through the ‘good offices’
function. Assistance was provided through the provision of medical
supplies, food, clothing, and cooking and consumable goods, often in
conjunction with NGOs, who would carry out the actual assistance
programmes. But this assistance did not provide UNHCR protection to
these groups or offer refugees any durable solutions. Assistance was
provided to the Chinese in Hong Kong and Macao, the Tibetans in India
and Nepal, and Indochinese refugees fleeing colonial France due to
independence struggles. However, these groups were not given prima facie
refugee recognition. Instead, they were assisted under the ‘good offices’
provisions because ‘eligibility for the status of refugee could not be
specifically determined’.63 However, as mentioned earlier, Schnyder was
concerned with the increasing gap between ‘old’ or Convention refugees
and the increasing number of ‘new’ or Statute refugees. Such a gap, he
felt, was untenable, particularly as the majority of the UNHCR’s work was
being conducted in developing countries.64 Unlike the majority of African
states though, Asian states had not signed onto the 1951 Convention or
developed regional refugee mechanisms. Although they could have opted
out of the time and geographic constraints in their accession to the 1951

62 Holborn, above n. 13, 657.

63 Holborn, ibid., 658.
64 UNHCR, Explanatory Note, above n. 1; see also, Loescher, above n. 10, ‘Chapter Five: Expansion
into Africa under Schnyder’.
718 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

Convention, Asian states still resisted overtures by the UNHCR during this
period.65
This, then, brings us to the third key issue underlying the UNHCR’s
advocacy of a Protocol, the potential emergence of regional refugee
mechanisms. The growing refugee problem in Africa led to the emergence
of a regional refugee instrument, the Organisation of African Unity (OAU)
Refugee Convention. This contained a broader refugee definition that took
into account the possibility of mass influx and generalised fears of violence.
However, Deputy High Commissioner Sadruddin Aga Khan spoke with
relief when the OAU decided that African states, though members of the
OAU Refugee Convention, still needed to accede to the 1951 Convention.
He declared that this demonstrated that the Convention had become
‘more universally recognized’66 – implying, of course, that it was not
before. However, Algeria and Nigeria both quickly pointed out that the
OAU Convention did not mitigate the limitations that existed in the 1951
Convention, its lack of applicability to Africa, and the need for it to be
reformed or replaced with a new instrument.67 So, although the UNHCR
was quick to applaud African states for taking the initiative in developing a
regional instrument, it was also determined to ensure that the region
continued to acknowledge the UNHCR Statute and the 1951 Convention
as the primary global instruments; regional instruments would be
secondary.68 However, the hostility of many African states to the 1951
Convention also confirmed the need for the UNHCR to start considering
the views of developing states more seriously and to recognise the different
needs of these new states.69 Likewise, the lack of accessions in Asia
demonstrated a need for the UNHCR to make the refugee instrument
relevant to Asian states.
While Africa’s political leaders were ‘eager to raise their status and improve
their image in the international political system’ through belonging to a
global refugee regime, the UNHCR was happy to ‘socialize new states to
accept the promotion of refugees norms as part of becoming law-abiding
members of the international community’.70 Many Asian states did not

65 UNHCR, Deputy High Commissioner to J. Asscher, Interoffice Memorandum: 1951 Convention,

21 Mar. 1956, Folio 535, 16/1/3 GEN, Series 1, Fonds 11. Overtures to individual Asian states were also
made regarding accession, so as to encourage other Asian states: UNHCR, Deputy High Commissioner
to Minister for Foreign Affairs of Afghanistan, Correspondence, 3 Nov. 1959, Folio 849, 16/1/3 GEN,
Series 1, Fonds 11, 2; UNHCR, Deputy High Commissioner to Prime Minister and Minister for
Foreign Affairs of Federation of Malaya, Correspondence, 4 Nov. 1959, Folio 852, 16/1/3 GEN, Series 1,
Fonds 11, 2.
66 UNGA, Report on the Fourteenth Session of the Executive Committee of the High Commissioner’s Programme,
Official Records, Twentieth Session, 25 Oct. – 2 Nov. 1965, 23.
67 Ibid., 24.
68 UNHCR, Visit of Dr Urrutia to Legal Division on 13 April 1966, 18 Apr. 1966, Folio 1602, 16/1/3
GEN, Series 1, Fonds 11, 2; Soguk, above n. 25, 174; Loescher, above n. 10, 125.
69 Loescher, ibid., 124-26.
70 Loescher, ibid., 105-6.
 The Shaping of the 1967 Protocol 719

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

adopt the same position, and this was reflected in the lack of Asian accessions
to the instruments in the 1960s.71 However, Asian states were not alone in
their reluctance to sign the 1951 Convention because the ‘new refugee
situations are, of course, no more covered by that dateline’.72 In 1965, several
Latin American states also expressed disquiet at the Convention, as they felt
it had ‘no application to their problems because of the dateline’.73 This
general level of discontent with international refugee law was more clearly
articulated once discussions about developing a new refugee instrument
began in earnest in 1965, and it is here that we see how important the political
context is in shaping the development of international refugee law.

4. The Bellagio Colloquium – creating the 1967 Protocol

In 1964, the High Commissioner, supported by the Belgian delegate on
the Executive Committee,74 proposed an initiative to resolve the situation
where the ‘High Commissioner deals with the “old” groups of refugees
on the basis of the Statue and assists new groups of refugees by virtue of
his good offices function … without necessarily having to examine their
eligibility under the Statute’.75 For, under the Convention, ‘hardly any of
the new groups of refugees can be considered as falling within its scope
owing to the dateline of 1 January 1951’.76 The key question, however,
was how the UNHCR and the Executive Committee could go about
removing the dateline. The Belgian delegate, who first raised the issue,
suggested a supplement to the Convention, such as a Protocol.77 However,
High Commissioner Schnyder argued that a number of states had
expressed a reluctance to ‘assume legal obligations with regard to new
refugee situations the nature and scope of which they cannot yet foresee’.78
This was not very different to the arguments leveled in 1951, when an
overriding concern about signing a binding obligation for a future number
of unknown refugee groups produced a legal definition of the refugee
that owed more to prudential calculations of national interest than

71 UNHCR, I. C. Jackson to The Representative, UNHCR Branch Office for Nepal, Accession by
Nepal to the 1951 Convention and the Protocol, 16 Sept. 1967, Folio 4741, 16/1/3 GEN, Series 1, Fonds 11;
UNHCR, W. K. McCoy to UNHCR Geneva, Memorandum: Malaysia – Attitude to Convention, 15 Mar.
1969, Folio 1829, 16/1/3 GEN, Series 1, Fonds 11; UNHCR, W. K. McCoy to UNHCR Geneva,
Memorandum: Singapore – Attitude to Convention, 24 Mar. 1969, Folio 1831, 16/1/3 GEN, Series 1, Fonds
11; UNHCR, W. K. McCoy to UNHCR Geneva, Memorandum: Indonesia – Attitude to Convention, 6 May
1969, Folio 1839, 16/1/3 GEN, Series 1, Fonds 11; UNHCR, E. Jahn to Mr Goundian, Asian-African
Committee, 6 Feb. 1970, Folio 1879, 16/1/3 GEN, Series 1, Fonds 11, 2.
72 UNHCR, High Commissioner Schnyder to Mr John Goormaghtigh, Correspondence, 20 July 1964,
Folio 16, 16/1/3 GEN, Series 1, Fonds 11, 1.
73 UNHCR, Visit of Dr Urrutia, above n. 68.
74 UNHCR, Correspondence, above n. 72, 2.
75 UNHCR, Explanatory Note, above n. 1, 2.
76 Ibid.
77 UNHCR, Correspondence, above n. 72, 2.
78 Ibid.
720 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

humanitarian sentiment.79 The UNHCR therefore proposed a colloquium
at which a number of legal experts representing their respective states
and regions would discuss the removal of the dateline and how such a
change would be attached to the 1951 Convention.
Thirteen legal experts attended the Colloquium on Legal Aspects of
Refugee Problems in Bellagio, Italy, from 21 to 28 April 1965. The countries
from which the participants originated were Algeria, Colombia, France,
Ghana, Hungary, India, Italy, Norway, Switzerland, United Kingdom,
United States and Yugoslavia (the thirteenth delegate was from the World
Council of Churches).80 What was notable about this colloquium at the
outset though was its members’ insistence that there would be no revision
of the Convention itself, 81 although it was:
(I)ncreasingly recognized that the refugee problem was universal in nature and of
indefinite duration and that, in this respect, the Convention was therefore no
longer adequate. An increasing number of refugees were not covered by its terms,
particularly as it was limited to persons who had become refugees as a result of
events occurring before 1 January 1951.82
By late 1966, the new High Commissioner, Sadruddin Aga Khan, went
so far as to declare a Protocol necessary because:
The Convention had led to an unfortunate discrimination among the different groups
of refugees, in particular with regard to the African refugees. Such discrimination
conflicted with the Statute of his Office and was contrary to the universal spirit of
the Convention itself.83
Regardless of the Convention’s noted limitations, the result of the
Colloquium was an agreement that a Protocol be attached to the 1951
Convention, removing the time and geographic constraints, and that
accession to the Protocol would bring automatic accession to all the other
articles in the 1951 Convention (therefore, separate accession to both
instruments was not required). This proposal was passed by the Executive
Committee of the High Commissioner’s Programme in November
1966.84 The draft Protocol was then submitted to the General Assembly,
after consideration by the Economic and Social Council. To all intents
and purposes, it appeared that developing states were finally receiving the
independent, universal refugee instrument they had long called for.

79 Mr Lawrence Finkelstein to Mr Boyd Compton, Memorandum: Proposed Colloquium on Legal Status of

Refugees, 21 Oct. 1964, Rockefeller Foundation Archives, above n. 33, 2.
80 UNGA, Colloquium, above n. 33, Rockefeller Foundation Archives, ibid.
81 UNGA, ibid., Rockefeller Foundation Archives, ibid., 2-4.
82 UNHCR, Memorandum, above n. 7, 2.
83 UNGA, above n. 17.
84 UNGA, Summary Records of the 141st to the 152nd Meetings held from 31 October to 8 November 1966,
Executive Committee of the High Commissioner’s Programme, Sixteenth Session, A/AC.96/SR.
141-152, 71.
 The Shaping of the 1967 Protocol 721

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

In principle, it appeared that the Protocol would have the effect of
universalizing the Convention and it was recognised as a separate legal
document to the 1951 Convention.85
On closer inspection, however, the Protocol did not fully address these
concerns and it ultimately failed to rectify the 1951 Convention’s inherent
limitations noted by Sadruddin in 1966.86 This judgment is based on two
principal factors. The first was the choice of a Protocol rather than a revised
Convention to replace the anachronistic 1951 Convention. Considering the
force and regularity with which states such as Nigeria, Algeria, and Asian
and Latin American states (see above) objected to the 1951 Convention, it is
not at all clear why a complete review of the 1951 Convention did not take
place. Second, the Protocol essentially universalised the Eurocentric
definition of a refugee and attendant support mechanisms, rather than
developing a system based on the type of refugee crises confronting Africa,
Asia and elsewhere. As a result, the Protocol failed to progress according to
the needs that High Commissioner Schnyder had identified when first
broaching the topic in 1964. Therefore, the outcome of the 1967 Protocol
was vastly different to the UNHCR’s intention. The final two sections of this
article will address this failure. First, it will look at the problems with selecting
a ‘Protocol’ as opposed to alternative routes for broadening the 1951
Convention. Second, it will look at the failure of the Protocol to broaden the
refugee definition. In the conclusion the relationship between politics and
law will be assessed, as well as how the power that dominates this relationship
has led to the story of the 1967 Protocol being forgotten and its outcome
being accepted as analogous to the UNHCR’s initial intention.

5. Why the Protocol?

The Protocol has been described as an ‘innovation in international law’;87
in contrast to the 1951 Convention, it was quickly developed and
promulgated in order to resolve growing complaints from developing
states about the UNHCR and the 1951 Convention. Within two years of
the Colloquium’s draft Protocol being presented to the UNHCR’s
Executive Committee, the Protocol had come into force.88 The question
remains, however, why was a Protocol decided upon and what benefits/
losses came from this approach?

85 However, Hathaway argues that the attempt to ‘universalise’ the 1951 Convention through the
1967 Protocol did not lead to a substantive review of the definition. Hathaway, above n. 16. Arguments
supporting the 1967 Protocol as being the hallmark of the 1951 Convention’s universalisation are:
Holborn, above n. 13, 179; Soguk, above n. 25, 173-174; Loescher, above n. 10, 124-125. Goodwin-
Gill expresses doubt in general about the necessity of the separate instrument: Goodwin-Gill, above
n. 6, 297.
86 UNGA, above n. 17; Hathaway, above n. 16.
87 Holborn, above n. 13, 181.
88 Holborn, ibid., 180.
722 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

It was agreed by the Bellagio Colloquium and the UNHCR that a
Protocol was the most effective way to remove the Convention’s time
constraints for a number of reasons. First, the time constraints needed to
be removed as swiftly as possible, so that refugee groups not covered at the
time would be ‘granted similar benefits by means of an international
instrument’.89 A legally binding instrument was necessary, but a new
Convention would ‘be too lengthy and cumbersome to beat the need for
urgency’.90 The Protocol would be the most effective means of avoiding:
(T)he traditional amendment process which would have required calling an
international conference of representatives of every government that was a party
to the Convention. Such a conference would have consumed a great deal of time,
would have opened up the entire Convention to possible amendments, and would
have meant that any amendment proposed had to be accepted by every contracting
party before it took effect.91
Secondly, a Protocol was seen as the way to go rather than an even
quicker approach (for example a resolution or recommendation) because
these more expedient alternatives would have no legal force and would
‘lose the support of the African states’.92 As noted earlier, African support
was considered essential in the face of the emerging regional approaches.93
The UNHCR also feared that a resolution would open a Pandora’s Box
and prompt the General Assembly to discuss the ‘whole substance of the
Convention’.94 To avoid this, the UNHCR suggested that only ‘a limited
number of states’95 should discuss an amendment to the Convention.
The Pandora’s Box fear was grounded in the fact that there remained a
large number of governments resistant to the idea of accepting increased
legal obligations, as they saw it, to an unforeseen number of potential
refugees. Moreover, the High Commissioner wanted to limit the discussion
to a small number of legal experts representing Europe, Africa, Asia,
Latin America and the United States to permit the development of the
core principles on this matter, which would ‘stimulate a concerted state
practice’.96 Third, the Protocol effectively removed the time constraints
from the Convention, while creating a new legal instrument that would
be open to states not yet party of the Convention.97 Fourth, the Protocol,
when applied to ‘new refugee situations should take the existing legal
structure fully into account and should therefore be built upon the 1951

89 UNGA, Colloquium, above n. 33, Rockefeller Foundation Archives, above n. 33, 2.

90 UNHCR, Memorandum, above n. 30, 2.
91 Holborn, above n. 13, 181.
92 UNHCR, Visit of Dr Urrutia, above n. 68.
93 Loescher, above n. 10, 124-25.
94 UNHCR, Visit of Dr Urrutia, above n. 68.
95 UNHCR, Correspondence, above n. 72, 2.
96 Ibid.
97 UNHCR, Memorandum, above n. 30, 2; UNHCR, Visit of Dr Urrutia, above n. 68.
 The Shaping of the 1967 Protocol 723

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

Convention’.98 In the words of the High Commissioner – the 1951
Convention was still the ‘most comprehensive instrument so far adopted
in this field’.99 Moreover, the 1951 Convention was already seen as
containing ‘a definition of the term “refugee” which is capable of
universal application’.100 However, the same memorandum raised a
concern about the 1951 Convention’s refugee definition:
(E)xperience, and more especially recent experience of new refugee situations in
Africa, has shown that certain States may not, for various reasons, be able to resort
to individual eligibility determination … [and] it might be of advantage if the
existing definition were supplemented by certain criteria for the prima facie
determination of such group eligibility.101
Nevertheless, the UNHCR rationalised that this ‘legal problem’ could not
be considered in the Protocol debate because it was more important that
‘the new international instrument should be both simple and generally
acceptable’.102 The consequences of this decision will be discussed in the
following section on the outcome of the 1967 Protocol. For now, it is
important to note that further issues pertaining to the 1951 Convention
were raised but were dismissed, mainly out of concern that this would lead
to the failure of the Protocol or an unending and unproductive discussion
about the Convention.103 Ultimately, prudential considerations took
precedence over wider questions about the continued relevance of the
Convention.
The Protocol achieved its prudential aim in that there was little debate
in the Executive Committee or the General Assembly about the need for
such an instrument. Many developing states, in particular, had been
pushing for a long time for the UNHCR instrument to become universal,
and Western states were not perturbed by the decolonised regions having
to deal with their own refugee problems, now that it was not the concern
of the former colonisers.104 The desire for an instrument more responsive
to global refugee problems was reflected in the speed with which the 1967
Protocol came into force: it was passed by the General Assembly on
16 December and came into force on 4 October 1967. The 1967 Protocol
gained twenty-seven ratifications in the following two years; by 1972 fifty-
two states were party to it. This was in contrast to the 1951 Convention,
which took ten years to achieve just twenty-seven ratifications.105 Did the
1967 Protocol meet the expectations of the states that it was mainly

98 UNHCR, Memorandum, ibid., 3.

99 Ibid.
100 Ibid.
101 Ibid., 5.
102 Ibid. Emphasis added.
103 UNHCR, Visit of Dr Urrutia, above n. 68; Holborn, above n. 13, 181.
104 Holborn, ibid., 180-186; Loescher, above n. 10, 122-26.
105 Holborn, ibid., 182.
724 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

intended for – developing states dealing with large refugee groups? This
question will be addressed in the next section.

6. The inadequacy of the 1967 Protocol

At the Executive Committee session in 1966, the Protocol was placed on
the agenda for discussion, to then be submitted to the UN General
Assembly through the Economic and Social Council so that ‘the Secretary-
General might be authorized to open the Protocol for accession by
Governments within the shortest possible time’.106 The UNHCR was
confident of accessions because it had already confirmed a number of
states’ support for widening the Convention.107 In addition, the only
states questioned about the introduction of the Protocol were those
already members of the 1951 Convention, and the state members of the
Executive Committee at that time. Therefore, the total number of
governments consulted was fifty-eight.108 Membership of the Executive
Committee at the Sixteenth Session comprised twenty-nine member
states: five from Africa (Algeria, Madagascar, Nigeria, Tunisia and the
United Republic of Tanzania); three from Latin America (Brazil,
Colombia and Venezuela); and four from the Middle East (Iran, Israel,
Lebanon and Turkey). The remainder included fourteen states from
Europe (Austria, Belgium, Denmark, Federal Republic of Germany,
France, Greece, Holy See, Italy, Netherlands, Norway, Sweden,
Switzerland, United Kingdom of Great Britain and Northern Ireland,
and Yugoslavia), Australia, Canada, and United States of America. Only
one state – China – was from Asia.
The general consensus within the Executive Committee, which had the
authority to decide on whether the Protocol would progress to ECOSOC,
was that the 1951 Convention be made ‘universally applicable to
refugees’.109 Therefore, the Committee agreed to the High Commissioner
submitting the draft Protocol to the General Assembly, with the proviso
that some governments intended to make ‘further comments or submit
amendments when the draft Protocol was examined by the Economic and
Social Council and by the General Assembly’.110
There were some dissenting voices but these were limited in the General
Assembly. The main criticism of the Protocol did not come to a head until
it was passed at the General Assembly. During the discussions on the draft
Protocol, two African states expressed discontent. Nigeria argued that the
106 UNGA, Report of the United Nations High Commissioner for Refugees Part Two: Report on the Sixteenth
Session of the Executive Committee of the High Commissioner’s Programme from 31 October - 8 November 1966,
Official Records of Twenty-First Session, Supplement No. 11 (A/6311/Rev.1), 19.
107 Holborn, above n. 13, 180.
108 UNGA, above n. 106, 18.
109 Ibid.
110 Ibid.
 The Shaping of the 1967 Protocol 725

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

Protocol did not ‘go far enough to solve the problems’.111
First, the
assistance provided to African refugees ‘did not even represent one tenth
of the already very small subsistence allowance given to refugees elsewhere
than in Africa’.112 Second, many African countries with ‘still under-
developed’ economies were shouldering a ‘crushing burden when the
problem had not been of their making’ (emphasis added).113 Nigeria argued that
African refugee problems had been caused by colonialism, and, in
particular, by the countries that had colonised Africa. They had a duty,
Nigeria argued, to contribute to the UNHCR programme and ‘acknowledge
that they had a moral obligation to alleviate as much as possible the
hardships for which they were responsible’.114 Uganda indicated that it
would support the Protocol, but, like Nigeria, suggested that ‘the text of
the draft was too general in nature and felt that it did not place the emphasis
where it belonged’.115 In particular, the economic burden caused by rising
refugee populations in Africa required more specific resolutions to deal
with problems and the 1951 Convention could not facilitate this.116
Other concerns about the 1967 Protocol expressed by developing states
in the General Assembly largely focused on the need to provide more
development assistance for refugee populated areas. There were also calls
for more legal assistance and protection and it was hoped that the Protocol
would fulfill this need. This emphasis on the need for the Protocol to deliver
assistance benefits drove its swift passage and led to little dissent. However,
a number of states expressed concern about the rapidity with which the
1967 Protocol was passed and with the lack of wide participation in the
drafting and development of the Protocol.
At the end of the debate in the General Assembly – and after the vote
had been finalised – delegates from the United Arab Republic (today,
Egypt and Syria), Afghanistan and Jamaica all expressed concerns about
the Protocol because ‘as an important United Nations treaty, it should have
been discussed and adopted in its entirety before states were asked to
accede to it’.117 Madagascar also ‘would have preferred the General
Assembly to consider the articles of the draft protocol in the light of
comments by States … but had voted in favor of the draft resolution as a
whole because if felt that the extension of the High Commissioner’s
mandate should not be delayed’.118 The Philippines had also voted in favor

111 UNGA, Official Records of the Third Committee, Twenty-First Session, 1449th Meeting, 6 Dec. 1966,
436.
112 Ibid., 437.
113 Ibid.
114 Ibid.
115 Ibid., 440.
116 Ibid.
117 UNGA, Official Records of the Third Committee, Twenty-First Session, 1450th Meeting, 7 Dec. 1966,
446.
118 Ibid., 447.
726 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

of the Protocol because ‘it felt that there should be no discrimination
between persons who had become refugees before 1 January 1951 and
those who had become refugees after that date’119 and because of the
urgency of the High Commissioner’s work. Contrary to the 1967 Protocol
representing an ‘innovation’ for international law, the Philippines expressed
the opposite view, it:
(W)ould have preferred the Committee to discuss in detail every article of the
Protocol … in the future, his delegation would insist that the articles of important
international instruments should be considered in detail and would not consider
the present case as a precedent.120
As discussed above, a UNHCR internal memo on the Bellagio Colloquium
noted the need for the Protocol to be an ‘independent instrument’121
from the 1951 Convention. It needed to address the global nature of the
refugee problem while embodying the principles of the 1951 Convention.
To some extent, the 1967 Protocol reflected this need. However, it did
not address the severe conditions that refugees and their hosts in developing
countries endured, particularly, when compared to conditions in developed
states. The Protocol was unable to address the problems confronting
many developing countries faced with mass influxes of refugees because
the Convention’s insistence upon individual refugee status determination
was carried into the Protocol. Such individual determinations were simply
not possible in cases of mass influx or other emergencies in developing
countries. There are a number of reasons for this. First, there is often a
lack of bureaucratised structures to handle such crises, and, second, the
military traditionally handles border breaches as security threats,
preventing access to refugee determination procedures (even if they do
exist).122 Moreover, in developing countries many refugees are forced to
flee due to a general fear of violence caused by armed conflicts, as
opposed to the specific fear of political persecution required by the 1951
Convention.123 The 1967 Protocol did not change this situation; there
was no ‘certain criteria for the prima facie determination of such group
eligibility’124 within either of the refugee legal instruments. This
perpetuated a problem for developing states, forced to deal with mass
refugee influxes, without the bureaucratic capacity to make individual
status determination.

119 Ibid., 446.

120 Ibid.
121 UNHCR, Memorandum, above n. 30, 2-4.
122 Chatelard, above n. 21.
123 Schmeidl, above n. 19; McGuiness, above n. 20.
124 UNHCR, Memorandum, above n. 30, 5.
 The Shaping of the 1967 Protocol 727

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

7. Conclusion
Understanding the development of the 1967 Protocol is important
because it, arguably, falls short of Commissioner Schnyder’s intention for
the UNHCR to be more equitable in its provision of protection and
assistance. The UNHCR and recently decolonized states wanted changes
to the 1951 Convention because they believed that it placed the greatest
burden on developing states near to the source of a crisis. Yet, instead,
we saw, as we did with the lack of non-European participation in drafting
the 1951 Convention, this unequal representation continued during the
discussion that framed the 1967 Protocol. This confirmed the perception
that many non-European concerns would continue to remain unaddressed.
The 1951 Convention and the 1967 Protocol were not primarily developed
to respond to Asian and African concerns about refugee problems. As
James Hathaway argues, the 1951 Convention was primarily concerned
with serving the political needs of the West.125 The 1967 Protocol, though
important because it removed the time and geographic constraints
contained in the Convention, prohibited alternative interpretations of
who a refugee was and how this status could be determined. What is
apparent from the drafting procedures of the 1967 Protocol is that
developing states had little role in the process of creating this instrument
but were, nonetheless, expected to adopt it. This reveals that the UNHCR
wanted to universalise the 1951 Convention to enable it to assist refugees
globally. It was clear, by the 1960s, that individuals from the developing
world would dominate the refugee population, yet the instrument
developed to serve this vulnerable group did not reflect their needs or
interests.
Realists would argue that this does not matter, because international
legal rules are largely irrelevant amongst the factors that shape the way
states behave and, as a result, states feel no need to legitimate their action
by reference to prevailing norms (including international legal norms).126
Therefore, the development of these legal rules was shaped by power, and
states recognise this fact by manipulating the laws to suit themselves.
However, what if law is not secondary to power and politics? Rather, if we
see law as a ‘powerful, rhetorical tool’127 that shapes but does not determine
political decisions, the realist view can be challenged in that legal arguments
were used to legitimate behaviour.

125 Hathaway, above n. 16.

126 For further discussion of this, see, F. V. Kratochwil, Rules, Norms, and Decisions: On the Conditions of
Practical and Legal Reasoning in International Relations and Domestic Affairs, Cambridge Studies in International
Relations (No. 2), (Cambridge University Press, 1989).
127 F. V. Kratochwil, ‘How do Norms Matter?’, in M. Byers, (ed.), The Role of Law in International
Politics, (Oxford: Oxford University Press, 2000), 35-68, 45.
728 Sara E. Davies

Downloaded from https://academic.oup.com/ijrl/article/19/4/703/1515319 by BIBLIOTECA DE LA UNIVERSITAAT DE BARCELONA user on 28 April 2022

This view is important in analyzing the outcome of the 1967 Protocol,
because what happened during the drafting of the Protocol was proof of
the rhetorical power that the 1951 Convention had already obtained for
itself in the minds of powerful states. In addition, the widespread acceptance
of the 1967 Protocol as being no more than its outcome – a simple removal
of time and geographic constraints – indicates how readily we accept the
rhetoric of law and its apparent separation from politics.
This article has demonstrated that the rhetorical power of the 1951
Convention has dominated our understanding of what the UNHCR was
attempting in the early 1960s, and its consequence lives on today. We see
the same problems with international refugee law that were foreseen in the
1960s – many developing states unable to develop the bureaucratic largesse
to respond to asylum seekers; the narrowness of the definition precluding
many fleeing violence, famine, disease and natural disasters from seeking
refugee status; and the continuing growth of refugee flows in developing
regions.
Was the adoption of the 1967 Protocol the moment when the 1951
Convention and UNHCR itself became truly universal? In a small and
superficial sense it was, but in a larger more practical sense it was not. The
abandonment of the time and geographical limitations on the definition
of refugee status marked an important step forward. However, this was the
sum of the Protocol’s contribution. A problem remains with the refugee
definition as its focus on political persecution has been at the cost of
assisting and protecting larger numbers of people fleeing generalised
situations of violence and abuse.
States expressed discontent at the speed with which the 1967 Protocol
was passed and with its failure to address these issues at the voting session
of the Protocol. Yet these same states expressed a reluctance to voice such
concerns before the Protocol was passed fearing its rejection and the failure
to remove the time and geographic constraints required to universalise the
Convention.128 As a result, the 1967 Protocol was a compromise between
the short-term humanitarian needs of refugees and the interests of states
in Europe, Africa and Asia, which effectively sidelined serious discussion
about many of the key sources of tension. In tracing the development of
the Protocol we are able to observe that in commending ourselves on the
creation of these instruments it is, perhaps, easy to forget that they were
promulgated at the cost of excluding alternative forms of refugee
experience. In assessing the Protocol, and not just dismissing it as a simple
but unnecessary expansion of the 1951 Convention, we are reminded of
how the UNHCR understood this and attempted, but failed, to remedy
this exclusion.

128 UNGA, above n. 117, 446-447.