The UPR is a process, initiated in 2007, in which all UN member states’ human rights performances are examined over a 4 ½ year period by other states. The reviewing states are largely but not exclusively sitting UN Human Rights Council members. It functions against the backdrop of the manifold complexities of international relations, differing worldviews, and the modus operandi of an international institution based on consensus. One analyst puts it this way:
10. See Mary Fletcher, Bahrain Coordinating Committee, Blogpost:
UNHRC President Lasserre Puts Kingdom of Bahrain on Notice do not Threaten Activists (26 May 2012), available at https://bahraincoordinat- ingcommittee.wordpress.com/2012/05/26/unhrc-president-lasserre-puts-kingdom-of-bahrain-on-notice-do-not-threaten-activists/.
»The principal UN human rights organ is not a tribunal of impartial judges, not an academy of specialists in human rights, nor a club of human rights activists. It is a political organ composed of States represented by governments that as such reflect the political forces of the world as it is.«11
The HRC in general, and the UPR in particular, are po-litically driven. As an NGO observer bluntly stated, »The HRC is not going to act like an Amnesty International or a Swedish government«.12 It is composed of UN member states seeking to benefit from belonging to the organiza-tion. As such, dynamics such as regional or affinity af-filiations (e.g. Commonwealth, Organisation of Islamic Cooperation) will continue to play a central role. And as peer review mechanisms are all by definition creations of the member states of the organization undertaking them; they all tend to be the products of consensus. This in turn means that they rely more on the carrot of positive reinforcements and inducements rather than the stick of punitive measures.
The UPR’s stated purpose is to promote and deepen re-spect for human rights through the provision of feedback to member states on their human rights performance. It is based on the concept of »continuous improvement«, which »focuses on incremental, constantly monitored steps, rather than great leaps forward«.13 A three-member HRC committee oversees the preparation and presentation of information regarding the State under Review (SuR) adherence to a range of human rights criteria. Each SuR presents a self-assessment of its hu-man rights record. Member states and NGOs comment on this, and issue recommendations. The government of the SuR has the choice to either accept or not these recommendations.
UNGA Resolution 60/251 governing the functioning of the UPR emphasized that the process should be coop-erative, constructive, non-confrontational and non-polit-icized. The process was also supposed to be inter-govern-mental in nature and UN member-driven, not be overly burdensome or long, be realistic and was not to absorb
11. Marc Bossuyt, »The New Human Rights Council: A First Appraisal«, Netherlands Quarterly of Human Rights, vol. 24, no. 4, 2006.
12. NGO representative interview with the author, November 9, 2011, Geneva.
13. Hilary Charlesworth and Emma Larking, eds. Human Rights and the Universal Periodic Review: Rituals and Ritualism, Cambridge University Press, Cambridge, 2015, 20.
a disproportionate amount of time, human and financial resources. The lack of punitive sanctions attached to the process was very likely a sine qua non for the support of many southern states, who feared that the UPR could turn into a one-sided mechanism for western states to criticize their human rights records, warranted or not. At the same time, however, the relevant institutional proto-cols stated that the process should be »action-oriented«
and »not diminish the Council’s capacity to respond to urgent human rights situations«.14
The fundamental, underlying question is to what extent is this a meaningful exercise, or, by contrast, a feel-good exercise in emptiness? Reviews of the first cycle of the UPR, which ended in 2012, were generally guardedly positive but emphasized the incipient nature of the mechanism.15 It is now, however, possible to generate data and perspectives regarding a) the extent of imple-mentation of first cycle recommendations, and b) dynam-ics of the recommendations process within the second round compared to the first round.
a) First Cycle Recommendations Implementation
In October 2014 the non-partisan and non-governmental organization UPR Info issued an analysis of the extent to which SuRs have been implementing recommendations by the mid-term point between their first and second as-sessments in the established 4-year cycle.16 This report is based on compliance data provided not only by SuR gov-ernments, but also from civil society organizations from 165 countries. It is of course extremely important to have had the data from CSOs since governments can be ex-pected to have an interest in presenting their compliance record in the most favorable light. The report determined that almost half (48 %) of the 11,527 recommendations for which UPR received submissions from government and non-government stakeholders had resulted in full or
14. (A/HRC/5/21, 2007).
15. See, for example Charlesworth and Larking, eds., op. cit.; Emma Hickey, »The UN’s Universal Periodic Review: Is it Adding Value and Im-proving the Human Rights Situation on the Ground? A Critical Evaluation of the First Cycle and Recommendations for Reform«, Vienna Journal on International Constitutional Law, https://www.icl-journal.com/download/
a671e91c60a30231e1067f41ba849986/ICL_Thesis_Vol_7_4_13.pdf;
and Dr. Purna Sen, ed., »Two Years of Universal Periodic Review: Lessons, Hopes and Expectations«, Commonwealth Secretariat, London, 2010.
16. Beyond Promises: The Impact of UPR on the Ground; UPR Info, Ge-neva, 2014. http://www.upr-info.org/sites/default/files/general-docu-ment/pdf/2014_beyond_promises.pdf.
partial action by the mid-term point of assessment.17 Of these recommendations 38 percent were determined to be fully implemented, while the remainder were partially implemented.18
The report indicated that 19 % of non-accepted recom-mendations were also implemented. This may be due to the fact that SuRs found that some of these recommen-dations which they had determined to be unrealizable were in fact possible to implement, at least partially. In addition, 55 percent of accepted recommendations were implemented fully or partially by the mid-term point. The East European (EEG) group of nations was determined to have been most effective at implementing reforms, while Asian states were deemed to have been least effective.
The report (p. 5) suggests that issues with the highest percentage of implementation dealt with HIV-Aids, hu-man trafficking, and people with disabilities. By contrast, the greatest percentages of non-implementation within categories were of recommendations pertaining to the freedom of movement, right to land, and the death pen-alty, possibly because in many political contexts these issues are the most sensitive and challenging to address.
In addition, research by this author et al has suggested that civil society organizations engage with and have an impact in the UPR’s functioning.19 A central finding of this research is that official UPR state recommenda-tions do in fact incorporate perspectives and themes contained in CSO recommendations, although often they are phrased or nuanced differently.20 While there is no proof of causation, it is therefore credible to infer that CSO perspectives and input do have an impact on state recommendations. At a minimum, it appears that states share interests reflected by the CSOs. In addition, in many respects CSO recommendations mirror several overall patterns, for example in the percentage of accept-ance rates, the regional distribution of recommending states, and the distribution of action categories. These
17. It is very likely that in most cases the compliance figure would be higher by the end of each country‹s 4 year cycle between reviews.
18. Note that the term »partial compliance« can cover a range of levels of compliance.
19. Edward McMahon et al. »Do CSO Recommendations Matter?« Frie-drich Ebert Stiftung International Policy Analysis, 2013. http://library.fes.
de/pdf-files/iez/10343.pdf.
20. Lawrence Moss, »Opportunities for Nongovernmental Organiza-tion Advocacy in the Universal Periodic Review Process at the UN Human Rights Council«, Journal of Human Rights Practice, vol. 2, #1, 1.
factors provide greater legitimacy to the UPR process, as it appears to be reflecting at least some broader societal perspectives and concerns; the process thus benefits from CSO input.21
These findings provide some preliminary indications that the UPR is having a positive effect on human rights pro-motion in a global, macro sense. And in many countries human rights organizations are seeking to advocate for compliance; anecdotally suggestions have been made that the UPR has served to give such groups »cover« and the standing with which to hold the government to its UPR-related promises. It is important, however, that this perspective be complemented by additional quantita-tive, qualitative and case study research that can deepen understanding of the extent to which SuR compliance with recommendations does or does not both reflect and result in meaningful human rights promotion and protection.
b) UPR Analysis Format
It is a challenging task to make sense of what is really happening in a vast and complex mechanism such as the UPR, especially given that the UN consists of 193 mem-ber states. 14 states are reviewed per session, of which there are three each year. A plethora of documentation is produced for each review. Numerous stakeholders are involved. The whole process has resulted in to date over 46,000 recommendations categorized according to 58 different issues. The analyst is thus faced with the task of finding ways to disaggregate this mass of data in ways that can make sense of it.
The method employed in this paper is to look for ag-gregated patterns in recommendations. Clues can be discerned from, for example, the number of recommen-dations made, whether they are accepted by the SuR, the issues that they address, and the extent to which they reflect geographic considerations such as North-South or other regionally-oriented dynamics. Regime typologies also pose an interesting question; do more democratic states make different use of the UPR than more authori-tarian regimes?
We have also determined that the way in which recom-mendations are phrased can be extremely revealing in
21. Edward McMahon et al., »Do CSO Recommendations Matter?« op cit., 11.
terms of the intent of the recommending state. Are rec-ommendations phrased in a »soft« way, which can make it easy for the SuR to accept the recommendation and later claim compliance? Or are they posed in more rigor-ous language, which requires specificity of action and accountability? Depending on the issue these dynamics may play out somewhat differently, but we suggest that given the large sample size of the data generated by the UPR, basic trends can become evident.
In order to provide an empirical basis for analyzing these questions we have developed an action category scale which groups recommendations made based on the verbs utilized in the recommendation language. A rating of 1 is for recommendations directed at non-SuR states, or calling upon the SuR to request technical assistance, or share information; a rating of 2 is for recommendations to continue or maintain existing efforts; a rating of 3 is for recommendations to consider change; a rating of 4 is for recommendations of general action (i.e. address, promote, strengthen, etc.); and a rating of 5 denotes recommendations calling for specific, tangible and verifi-able actions.22
This approach provides a qualitative sense of the types of recommendations being made. In general Categories 1, 2 and 4 are more acceptable to SuRs and are easier to implement. Category 1 recommendations are addressed to other states to provide support and assistance. Cat-egory 2 recommendations simply require continuation of existing activities. The generality inherent in Category 4 recommendations means that the SuRs in question have considerable leeway in determining what constitutes suc-cessful fulfilment of the recommendation. Category 3 recommendations, by contrast, deal with more sensitive or unpalatable issues while Category 5 issues contain specific and verifiable required actions which more of-ten result in rejection. The UPR Info report, for example, states that the recommendations associated with the most specific actions (Category 5) received the lowest rates of implementation, while the recommendations emphasizing continuity of action, or actions of a gen-eral nature – which makes it easier for SuRs to define and thus assert compliance with – had higher levels of implementation.
22. More information on the action category scale is available at Edward McMahon (2012). The Universal Periodic Review: A Work in Progress.
Friedrich-Ebert-Stiftung, http://library.fes.de/pdf-files/bueros/genf/09297.
pdf and www.upr-info.org/IMG/pdf/Database_Action_Category.pdf.
Using this framework we prepared a report published in 2012 entitled The Universal Periodic Review: A Work in Progress, which discussed the development of the UPR.23 It examined how the mechanism has functioned, and what dynamics had emerged in how the mechanism is used. These included analysis of the types of recommen-dations, who made them, to whom they were directed, and to what extent they were accepted by the States under Review (SuRs). Analysis of this data from the first cycle revealed the following trends:
The mechanism was well-used; almost all states par-ticipated at a high level and the number of recommen-dations appeared (at the time) to have plateaued at just over 2000 recommendations per session.
Slightly over 2/3rds of all recommendations were ac-cepted, a percentage which increased moderately over the life of the first cycle. Acceptance rates were lower for more specific and action-oriented recommendations, and for those which requested States under Review (SuRs) to consider taking those types of recommendations.
Three-quarters of all recommendations were action-oriented, although somewhat over half of these were of a vague and/or general nature.
While recommendations were distributed regionally in proportion to UN membership, a large plurality of rec-ommendations were made by the Western European and Other (WEOG) group.
African and Asian states were more likely to make softer recommendations (i.e. to continue what an SuR was already doing, or to take a general action) while WEOG states were relatively more likely to issue specific action-oriented recommendations.
The most common recommendations addressed is-sues related to international instruments, women’s and children’s rights, torture and the administration of justice.
While it was difficult to aggregate data on Economic, Social Cultural as compared to Civil and Political Rights, the former appeared to have a higher acceptance rate.
More democratic states tended to make more action-oriented recommendations. This was true across regions.
We now examine sessions 13 through 22 of the Second Cycle as compared to the first cycle.24 We note that the totality of the second cycle is not represented here as the last session will be in November 2016, so we do not have
23. ibid.
24. There are a total of 26 sessions in the first two cycles.
the data for the last four of the 14 sessions in the Sec-ond Cycle. However, the number of recommendations that we do analyze is modestly greater than the total number of recommendations in the entire first cycle, and as it represents over 70 % of the session of the second cycle, and contains a statistically sufficient sample size to identify major trends.
The following data represent findings from this research.
3.1 Number of Recommendations
Graph 1: Total Number of Recommendations, Cycles 1 & 2
It is not surprising that the overall numbers of recom-mendations have increased between the two cycles. This is likely a function of the increasing familiarity and under-standing of the UPR process on the part of both mem-ber states and civil society organizations, which provide suggestions for recommendations to and transparency in assessing state adherence to accepted recommenda-tions. It is also an initial, prima facie suggestive indica-tor of interest and willingness on the part of member states to utilize this mechanism, although this does not of course necessarily reflect the utility or efficacy of the recommendations.
While in the early sessions of Cycle 1 member states tended to be tentative and exploratory in their approach to the UPR, the increase in recommendations was also evident when comparing the later sessions of Cycle 1 with Cycle 2. The overall number of recommendations per country (180) in Cycle 2 increased compared to the last four sessions of the first cycle (140). The increase continued within Cycle 2, as the number of recommen-dations per country averaged 153 for the first two Ses-sions (13 and 14) but increased to an average of 197 by sessions 21 and 22. We can speculate that the reasons
for the continued increase may be simply continued interest in the UPR. Also, they could reflect the reintro-duction of unfulfilled recommendations from Cycle 1, in addition to new recommendations. Recommendations from all regions increased markedly in both absolute and relative terms with the exception of WEOG, whose share of recommendations decreased for 41 % to 30 %.
3.2 Responses to Recommendations
15,636
18,893
5,720 6,335
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000
Number of Recommendations
Graph 2: Responses to recommendations, Cycles 1 & 2
Accepted Noted
Cycle 1 Cycle 2
The overall percentage of accepted responses has in-creased slightly, from 73 % to 75 %.25 This demonstrates a consistent rate of acceptance and noting of recom-mendations across both cycles and could be a predictor of future acceptance behavior. The fact that about three-quarters of all recommendations made are continuing to be accepted is important. Taken in conjunction with the fact that no state has refused to participate as an SuR, it suggests that states are seeking to be cooperative with the UPR.
25. In Cycle 1 non-accepted recommendations had been disaggregated according to the type of SuR response i.e. refusal, a general response, or no response at all. Given that they all reflect the same basic response – that of not accepting the recommendation, beginning with the second cycle the HRC decided to aggregate them into a single, »Noted«, cat-egory.
3.3 Breakdown of Acceptances by Region
29%
29%
13%
16%
13%
Graph 3: Percent of Total Accepted Recommendations by Regional Group SuR, Cycle 2
Africa Asia EEG GRULAC WEOG
The breakdown of member states in the UN General Assembly by geographic region is as follows: Africa 28 %, Asia 28 %, GRULAC 17 %, WEOG 15 % and EEG 12 %. The proportionality of accepted recommendations depicted above therefore mirrors almost exactly the UN membership by region. Also, Cycle 1 and 2 acceptance rates for the African, EEG and WEOG regions remained basically the same. By contrast, acceptance rates for Asia declined modestly from 76 % to 69 % while those for that Latin American states (GRULAC) were the exact op-posite of Asia’s, rising from 69 % to 77 %.
3.4 Breakdown by Action Category
Cycle 1 Cycle 2 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Graph 4: Breakdown by Action Category, Cycles 1 and 2
Category 1 Category 2 Category 3 Category 4 Category 5
The distribution by action recommendation category has remained very constant, with slightly less than 40 % of recommendations falling into Category 4 of general action recommendations, followed by about one-third being Category 5 of specific actions, and smaller percent-ages being accorded to the other three categories.26 The most notable difference between the two cycles is the increase in Category 2 recommendations, asking that the SuR »continue« to undertake particular actions, from 14 %, to 19 %. This is intuitively logical as at least some reviewing states pay attention to whether or not states have undertaken or completed actions to which they had agreed in the first cycle. Those recommendations which
The distribution by action recommendation category has remained very constant, with slightly less than 40 % of recommendations falling into Category 4 of general action recommendations, followed by about one-third being Category 5 of specific actions, and smaller percent-ages being accorded to the other three categories.26 The most notable difference between the two cycles is the increase in Category 2 recommendations, asking that the SuR »continue« to undertake particular actions, from 14 %, to 19 %. This is intuitively logical as at least some reviewing states pay attention to whether or not states have undertaken or completed actions to which they had agreed in the first cycle. Those recommendations which