search for




 

Power, Politicization, and Network Positions: Explaining State Participation in the UPR
The Korean Journal of International Studies 16-3 (December 2018), 335-65
Published online December 31, 2018
© 2018 The Korean Association of International Studies.

Su Hyen Bae [Bio-Data]
Received September 30, 2018; Revised December 13, 2018; Accepted December 18, 2018.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
The Universal Periodic Review (UPR) ensures formal equality among participating member states. However, previous literature emphasizes the interference of state interest and politics in undermining the universal peer evaluation mechanism. In this article, I argue that while the UPR shows certain bias in state behavior for providing recommendations, the UPR otherwise functions according to its purpose of condemning human rights violations. I find that member states’ human rights index scores and the level of democracy correlate with the number of recommendations received and the betweenness centrality measures. First, I apply social network analysis (SNA) on state interaction in the UPR literature enabling inter-network comparisons with international trade relations, military dispute, and alliance relationships. The QAP analyses depict that the UPR network has a low association with the military dispute network and the alliance network. Second, individual level analyses demonstrate that states with higher national capabilities and a greater amount of trade exports are more likely to provide recommendations. Nonetheless, higher human rights index levels lead to more recommendation providing activity while smaller in magnitude. However, the amount of recommendations received by a state suggests that states with low human rights records and low levels of democracy receive more recommendations. Furthermore, the betweenness centrality measures highly correlate with the human rights index and the level of democracy implying that the general standard of human rights influences the degree of state centrality in the UPR network. This study acknowledges the presence of politicization among states in providing recommendations, but also ensures that the UPR is shaming states according to its main purpose in criticizing the human rights violations of non-compliers.
Keywords : universal periodic review, social network analysis, human rights, politicization, betweenness centrality
INTRODUCTION

The Universal Periodic Review (UPR) provides a public forum for states to discuss human rights practices occurring in other states. All 193 United Nations (UN) member states participate in the peer-review mechanism providing recommendations to the state-under-review (SuR) and fulfilling its duty in being the SuR. However, each SuR receives varying numbers of recommendations from other states. In turn, states provide different numbers of recommendations to each SuR. In answering what accounts for such variation in state interaction patterns, previous works utilize a dyad level analysis (Hong 2018; Terman & Voeten 2018). Building upon past literature, this article employs social network analysis (SNA) to understand state interactions from an institutional perspective.


The UPR’s universality is one of its most significant characteristics that differentiate the state-to-state accountability process from the previous human rights institution, the Commission on Human Rights (CHR). The CHR was accused of selectivity, politicization, and double standards in scrutinizing state human rights violations (Landolt 2013; Carraro 2017). This culminated into pressure for a universal and equal process leading to the UPR. Nonetheless, politicization remains an unresolved issue. As Freedman and Houghton state, international institutions composed of state actors may not be able to escape politicization: “[p]olitical objectives will always be involved where a body or organization is intergovernmental…” (2017, 753).


How is the UPR influenced by politics and to what extent? This article tackles the presented question by offering a structural measure and delving into state positions. State interactions in intergovernmental organizations (IGOs) are commonly characterized as shaming and naming based on the reputation cost related to reneging from international norms. A human rights mechanism true to its peer-review purposes would engender interaction where states harshly criticize their allies and economic trade partners to an equal degree with states that they do not have strategic ties to. However, interdependence in complex international relations embed states in various networks. Therefore, state interactions in the UPR may not be solely caused by human rights mechanisms, but also the state relationships in other international networks.


Previous literature on the UPR anatomize the peer-review model and offer confounding studies on the extent of its politicization. Compared to the CHR, the UPR appears to facilitate the cooperation of authoritarian states, albeit encouraging their abuse of the system (Chow 2017). Nonetheless, when testing the UPR’s politicization at the individual state level, results convey a high influence of state decisions on whom to shame and whom to endorse based on political and economic factors (Schaefer & Groves 2010; Terman & Voeten 2018).


Thus, this research contributes to the debate on politicization by comparing the UPR network with other international networks in examining the degree to which interactions in the human rights mechanism resemble ties in economic and military alliance structures. This research is among the first to interpret the UPR’s interaction data as a relational social network data. Furthermore, structural comparisons enable a clearer understanding of the UPR’s politicization complementing conceptual definitions by providing quantitative measures.


From state interactions, this article creates an interaction network and derives the betweenness centrality that measures a state’s position within the UPR topology. Understanding the state’s social power through their structural positions in the network offers a unique perspective of the mechanism. The approach argues that interactions in the UPR engender a hierarchical structure of its own, deferring social power to states that have a higher centrality measure or central position among the community of states. In turn, the article attempts to answer why state centrality measures vary and the implications of the positional variation in the UPR concerning politicization.


I approach the issue in three steps. First, this article performs a network comparison analysis. Results demonstrate that the UPR network has a statistically significant correlation with the other international networks, but the extent of the networks’ similarity is low. Network comparisons are limited in its implications for they show associations not causal relations at the structural level. Hence, a second analysis is conducted to observe the associations among national power variables and the human rights index to explain individual state level of participation. The results demonstrate, conditionally, that active recommendation providers tend to have more national capacities. On the other hand, active recommendation receivers tend to have a relationship with low human rights records and low levels of democracy. In the third part of the analysis, I find factors influencing states’ central positions in the network. The betweenness centrality, a measure of power within the network for controlling the flow of information, correlates with higher human rights index scores and lower levels of democracy. In conclusion, this article confirms the presence of politicization among recommendation providers. Results reveal that the UPR rightfully shames states according to the states’ human rights records and that a state’s human rights record determines positions in the UPR interaction network.

NETWORKS OF IGO MEMBERSHIPS, TRADE, AND CONFLICT

Theoretically, this article begins with the assumptions from structural realism. A vast literature supports the liberalist perspective of IGOs in promoting peace and trade. However, such connectedness could influence IGOs reversely, which may cause politicization. Therefore, whereas past works focused on the extraneous effects of IGOs, this article perceives the other causal direction in looking for the effects of external factors that influence the IGO’s internal mechanisms. This article follows Hafner-Burton and Montgomery’s (2006; 2012) structural realism approach on conceptualizing state’s social positions within the IGOs.


Applications of networks as a theoretical, measurement, and inferential tool reflect international politics’ interactive and interdependent state relationships (Dorussen et al. 2016, 283). Interdependence is also measured through direct relationships between pairs of actors. However, interdependence does not terminate at the dyad, since one dyadic pair is likely to be influenced by the actions of another dyadic pair (Gartzke & Gleditsch 2008; Poast 2010; Dorussen et al. 2016). In understanding such complex interdependencies, previous contributions connecting network analysis and international relations have focused on the effect of networked interdependencies on conflict and cooperation.


Past research on shared IGO memberships, trade dependencies, and conflict emphasizes that interdependence decreases conflict. Russett and Oneal (2001) find evidence for the liberal peace argument by focusing on the cost of force over gains from trade. In support, Hegre and colleagues (2010) argue for the hypothesis that trade dependencies reduce the likelihood of state conflict. On IGOs and peace, Boehmer, Gartzke, and Nordstrom (2004) show how institutionalized IGOs reduce probabilities of militarized disputes (MIDs), but that other IGOs have a small influence on conflict.


Incorporating network concepts, Ingram, Robinson, and Busch (2005) demonstrate that trade between two countries increase substantially when the strength of their IGO ties increase. In relation, the result contributes theoretically to verifying the interdependence between different levels of institutions. Dorussen and Ward (2010) claim that Kantian peace variables have a significant negative effect on the likelihood of conflict. Lupu and Traag (2012) also support the Kantian peace theory demonstrating that trade flows within trading communities significantly reduce possibilities of conflict. Haim (2016) finds that the influence of dyadic alliances on trade is reduced after accounting for indirect relationships, but joint membership in an alliance community predicts an increase in trade.


However, recent works demonstrate that network dependencies increase possibilities for conflict leading to a re-evaluation on conventional interdependence claims. Hafner-Burton and Montgomery (2006) argue that IGO memberships create relative positions altering social power which explains conflict between states. Hafner-Burton and Montgomery (2012) also posit that trade interdependence produced through preferential trade agreements create asymmetrical relationships that exacerbate conflict. Structural equivalence increases militarized disputes (MIDs), while large prestige disparities decrease conflict.


Primarily, this article seeks to examine whether economic-military relationships explain UPR interaction patterns. Previous findings on the UPR pointed out that states condemn friends and allies less harshly in order to maintain strategic ties. Same geographical region, alliance, and aid relationships affect state participation (Terman & Voeten 2018). Hence, material interests and strategic concerns influence state relationships at the dyadic level. Whether the hypothesis could be supported at the structural level is one of the main aims of this article. Structural realism enables capturing state positions within the UPR as derived from material power. As such, the betweenness centrality is evaluated in terms of state national capacity in this article. On the other hand, the application of network analysis adds another perspective on power as stemming from the network position. Power is constituted from the position itself, not solely from individual attributes (Hafner-Burton & Montgomery 2006, 570).

RELATIONAL MECHANISM OF THE UPR

Ties among states are political in nature. It is difficult to sanitize the “most elaborate multilateral human rights process” in terms of political relationships (Terman & Voeten 2018, 2). However, if the dialogue and interactions are affected mainly by ulterior motives, this would pose a serious threat to the peer-review process and faith in international organization’s capacity to contain and curb human rights violations.


The UPR, established in 2008 under the UN Human Rights Council (HRC), promotes a universal approach and equal treatment by achieving voluntary participation of all 195 countries (including Palestine and Holy See) in the UN. The review process begins with the SuR turning in a national report to the UPR Working Group consisting of forty-seven member states. Then interactive dialogue takes place after the self-evaluation wherein all participating UPR states may choose to deliver a recommendation. In the first cycle, thirty-nine states did not make any recommendations (UPR-Info 2018). The SuR then responds which recommendations it will accept. A total of fifty-six issue categories are topics of discussion; the top five issues raised in the first cycle were international instruments, women’s rights, rights of the child, torture, and justice (UPR-Info 2018). After the review, randomly chosen three states compile an outcome report. Within 4.5 years, the SuR has the commitment to implement the accepted recommendations.


Some recognize the non-confrontational dialogue centered, universal appeal of the UPR, mentioning its advancements in garnering the participation of previous non-compliant states and focusing on universal issues not contained to cultural relativism (Gaer 2007; Redondo 2012; Soh & Kim 2014; Milewicz & Goodin 2016). Redondo (2012) asserts that the cooperative approach provides a unique context where public promises breed binding obligations leading to higher effectiveness. Hong (2018) finds a positive relationship between commitment to human rights treaties and higher monitoring performances in the UPR.


Nevertheless, the UPR has been subject to concerns on its vulnerability to political influences (McMahon & Ascherio 2012; Cochrane & McNeilly 2013; Freedman & Houghton 2017). By examining dyadic relationships, Terman & Voeten (2018) confirm the presence of a politicized, selective process where states are more lenient towards strategic relationships and that their recommendations are accepted more often than identical recommendations from states with fewer strategic ties. Carraro (2017) investigates that the perceived level of politicization in the UPR is high among member state representatives, but also brings about a higher responsibility to comply with commitments. Chow (2017) mentions that authoritarian states, such as North Korea, utilize UPR’s dialogue mechanism to justify its non-compliance, frame its human rights record in a positive light, and to condemn its political rival’s human rights violations. Accordingly, states with weak human rights records praised North Korea, whereas states with strong human rights records were more critical. Continuation of politicization may lead to reducing the mechanism to a series of empty rituals (Kälin 2014).


Only a modicum of literature empirically examine politicization, the pursuit of political objectives unrelated to human rights, a central interest in human rights discourse. The politicization of IGOs is a relatively unexplored topic (Reich 2005, 782), wherein literature focuses on the World Trade Organization (WTO), World Bank, or the UN specialized agencies in general (Brown 2011; Ghebali 1985; Lyons et al. 1977; Reich 2005). There are various definitions of politicization1 ; to combine the definitions from Reich (2005) and Ghebali (1985), the general idea of politicization refers to a situation where states advance unrelated, extraneous, objectives to the function of the UPR for politically motivated reasons. The UN’s specialized agencies seek to promote nonpolitical goals, so embroilment in political conflicts erode the authority and credibility of international institutions (Reich 2005, 782-3). However, turning to Lyons et al. (1977), politicization could be understood not only in pejorative terms, such as organizational defect, but also as a bargaining process wherein developing countries may promote a significant issue that extends beyond the technical goals of the organization (88). That is, UN specialized agencies may incorporate politicization to integrate dissatisfied powers into the international system (Lyons et al. 1977, 88).2 Such a view of politicization accords with Freedman’s assessment of the HRC that weaker states have politicized the institution more frequently than stronger states by bloc voting or forming alliances (2013, 120).


Within the HRC and UPR literature, Freedman (2013) and Carraro (2017) provide an institution-specific overview of politicization. Freedman (2013) acknowledges that politicization will always occur to some degree within international organizations that consist of members states with different legal and social structures, but problematizes the HRC’s politicization for its regionalism, or rather alliance formation of political groups with similar economic and social factors (119). In relation, although focusing on the variety of HRC activity, not only the UPR, Freedman and Houghton (2017) argues states advance political objectives and shields allies by using group tactics (753-4). In this vein, examining the relational, network structure of the UPR contributes to previous works’ concerns on nation blocs and alliances.


Carraro (2017) develops three indicators of politicization specifically within the UPR, regarding human rights reviews as country bias, issue bias, and instrumental use of cultural relativism. Country bias refers to differential treatment among countries that have comparable human rights performance. Issue bias is whether some issues are systematically given more attention for political reasons. Instrumental use of cultural relativism deals with the extent of the universality of a certain issue and how countries may employ cultural relativism to justify their human rights transgressions in the review process (Carraro 2017).


This article focuses on the country bias aspect of politicization rooted in national power and social relationships. All else equal, do states with more power and connections provide more criticisms toward other states? The argument is aligns with the position of structural realism that explains sources of state conflict formed within the confines of the international power politic environment (Hafner-Burton & Montgomery 2006). This perspective enables us to understand a state’s position within the international network as an emergent property of power influencing other state’s foreign policy behaviors (Hafner-Burton & Montgomery 2006). Within intergovernmental organizations, states are motivated by security and economic interests, and less by interest in promoting human rights itself. Hence, the concern is whether states undergo the risk of harming allies’ or dependent actors’ reputation while harshly condemning political rivals’ behaviors, mostly attributed to extraneous motives (Lebovic & Voeten 2006).


I construe the hypotheses based on the literature mentioned above on politicization and structural realism theory emphasizing state material capabilities. The network level hypotheses test the degree to which the UPR relationships correlate with other international structural relationships. I hypothesize that the UPR network has a significant overlap with economic trade, military dispute, and alliance networks. The three networks are leading indicators of national power and international connections. For example, if there is a high correlation between the UPR network and the economic trade network, then it signifies that interactions in the UPR reflect states’ economic interests and relationships with trade partners. Patterns of international trade and alliances help determine country incentives in defending or attacking one another (Jackson & Nei 2015).



Hypothesis 1: There is a significant association between the UPR network and international networks on economic and military ties.


As the network analysis determines the degree of correlation, the hypothesis centers on predicting whether there is an association or not. A close trade relationship, participation in military disputes, and close military alliance may not predict high or low participation in the UPR. That is, a country with a close trade relationship with another country may decide to provide a compliment to the other country. Unless the research design takes into account the content of the recommendations, it is difficult to distinguish whether the interaction was based on amity or enmity. Nonetheless, a significant association between the UPR network and the international trade network or the military network may verify the degree of overlap among international structures. It will provide news ways for interpreting interdependence among state relationships for the human rights mechanism. A state’s central position in the UPR network may overlap with a state’s central position in the international trade network.


Upon conducting the network level analysis, this article centers on the individual level data to complement the structural results. At the individual level, the network variables on trade, military disputes, and alliances are scrutinized. Because the individual level data also includes directions separate hypotheses have been introduced. The in-degree is equivalent to the number of recommendations received by a state. The out-degree is equivalent to the number of recommendations provided by a state. . Additionally, the Composite Index of National Capabilities Score (CINC) is utilized as one of the main independent variables indicating national power.



Hypothesis 2 (In-degree): The smaller the national power, the more likely it is for the state to receive recommendations.


In relation, it is expected that the lower the CINC index, the more likely it is for a state to receive recommendations. States with low national capabilities may be more vulnerable to attacks on human rights violations due to the lack of resources to adequately adopt human rights norms, and also because those states lack material power, which is less of a threat to countries wishing to provide criticism. Next, the smaller the amount of trade involved, the more likely it is for the state to receive recommendations. The trade variable is a measure of state exports to the entire world; the amount of export variable was utilized instead of the amount of import because the two variables high multicollinearity (correlation of 0.889)number of imports was statistically insignificant. Further tests also proved that the amount of imports have no significant relationship with state activity in the UPR.


Research on trade and alliances demonstrate that countries with a high level of trade with one another are less likely to be involved in conflicts (Jackson & Nei 2015). High export levels could indicate a level of trade activity, representing a state’s economic power in the global structure. States have fewer incentives and higher costs to criticize economically powerful states. Regarding military power and relationships, it is hypothesized that the greater the number of initiated military disputes, the more likely it is for the state to receive recommendations. Finally, the fewer the number of alliances, the more likely it is for the state to receive recommendations.



Hypothesis 3 (Out-degree): The greater the national power, the more likely it is for the state to provide recommendations.


The related sub-hypotheses are as follows. First, the higher the CINC index, the more likely it is for the state to provide recommendations. Second, the smaller the amount of trade involved, the more likely it is for the state to provide recommendations. Third, the greater the number of having been a target of military disputes, the more likely it is for the state to provide recommendations. Fourth, the greater the number of alliances, the less likely it is for the state to provide recommendations.


The sub-hypotheses follow the logic that states with power have a lower barrier to criticizing other states, which align with results from Terman and Voeten (2018) that states act more lenient towards their strategic partners. However, Terman and Voeten (2018) utilize a dyad level research design examining the interaction between each pair of states. Results shows the positive effect of geopolitical affinity, aid donorship, and military alliance in increasing participation in the UPR, while decreasing the severity of recommendations (Terman & Voeten 2018, 14). The dyad model provides a detailed result on state participation and the severity of comment based on the reviewers ties with the SuR. Complementing previous dyadic analyses, this article offers an aggregated individual level result focusing on state characteristics in determining participation



Hypothesis 4: The greater the national power, the more likely it is for the state to have a higher betweenness centrality.


Betweenness centrality is a network property measured for all individual nodes based on their position in the UPR network. It does not incorporate ties’ directions. Betweenness centrality demonstrates the extent to which the network is dependent on a specific node in maintaining connectedness (Hafner-Burton et al. 2009, 564). Utilizing the measure, nodes are described as mediators or gatekeepers because the betweenness centrality shows the amount of information flow controlled by the single node (Hafner-Burton et al. 2009, 565). As one of the main centralities, this hypothesis seeks to explain network position of states concerning their material power. The actors with high betweenness centrality bridge disconnected groups of actors within a network (Murdie & Davis 2012a, 192).


Networks have a socializing effect on members where human rights norms could diffuse through peer pressure (Finnemore & Sikkink 2001; Murdie & Davis 2012b). Yet, previous research acknowledges the ongoing politicization of UPR due to states’ selectivity in the provision of recommendations. In light of these arguments, I focus on depicting relations of all nations toward one another, which is vital in understanding the general political dynamics. While the UPR prima facie ensures equal opportunities in the review process, in reality, states enjoy differing degrees of actual equality since a discrepancy exists in the number of countries participating in the interactive dialogue with the SuR (Smith 2011).

RESEARCH DESIGN

The analysis proceeds in three stages. First, I compare the first cycle UPR network with the international trade network, military dispute network, and alliance network. Network regression analyses are performed to reveal the degree to which the UPR network is correlated with other international networks. Second, I conduct multiple linear regression analyses for node level data on in-degree and out-degree to find out the key variables that affect variation in the UPR activity and performance. This step is to understand the result of the network comparison more deeply at the individual state level. If the UPR network demonstrates a high correlation with the other networks, an individual level statistical model will enable us to determine precisely what the main influences are. Third, a node level network attribute, the betweenness centrality, is further analyzed with multiple regression analysis. The purpose here is to determine why some states acquire the betweenness centrality position. In other words, the question is, “do the same variables that affect UPR activity at the individual level, also help explain network structural positions held by individual states?”


To test the hypotheses, I created a new individual level and directed network dataset. Recommendation data was obtained from UPR Info on the first (2008-2012) and second cycles (2012-2016) consisting of 26 sessions from 2008 to 2016. The second cycle data is utilized to compare network level features with the first cycle. However, due to the lack of data for explanatory variables from 2012 to 2016, other analyses and statistic models comprise only of the first cycle data. The UPR Info database contains all the recommendations until the twenty seventh session aligned with the information on the state offering the recommendation (Reviewer), and the state receiving the recommendation (SuR).



DEPENDENT VARIABLES NETWORK LEVEL


This research is concerned with both individual level design and the network level design. Network level descriptive statistics for both first and second cycle UPRs are examined. The most essential network here is the directed and valued first cycle UPR network, which serves as the basis for the in-degree, out-degree, and betweenness centrality measures. This article operationalizes ties between states as the frequency of interaction in the UPR, with the strength of the tie equal to the total number of recommendations received or given. Directed signifies that a node’s connection to other nodes consists of an in and out direction. Thereby the in-degree of a node is the number of recommendations received, and the out-degree is the number of recommendations given, and the network is composed of both data for each of the nodes. It is also asymmetrical since node A may have received a recommendation from node B, but it is possible to withhold from reciprocating with a recommendation. Valued signifies that the frequency of interactions is calculated. For example, if node A received two recommendations from node B and three recommendations from node C, then ties with B and C are weighted with two and three. On the other hand, binary networks did not consider the frequencies but counted as one if an interaction existed and zero otherwise, so that node A’s ties with B and C would be marked as simply one.


The average is the average value of ties, which is sum divided by observation (20890/35532). This also signifies the proportion of possible ties that are present. The probability that any given tie between two random nodes in the first cycle directed/valued network is 58.8% and in the second cycle directed/valued network 96%. In comparison to the first cycle, the second cycle UPR demonstrates a higher level of connectivity among the nodes with the possible ties being present. The coefficient of variation is 215.5 (variance multiplied by 100), which suggests a high level of variation.


The average degree captures the sum of an actor’s in-degree and out-degree; an average of around 35 ties are present for each actor. Dyad reciprocity is the proportion of dyads that have reciprocated ties, which reveals a low level of reciprocity in the UPR network. The ‘clustering coefficient’ measures the degree to which a node’s local ties are closely knit between zero and one. The clustering coefficient for the network is high. Because it is difficult to infer about the network’s bias through analysis of the network itself, comparing the UPR network with other networks is necessary. The first cycle directed/valued network and first cycle directed/binary network are utilized to perform network regressions with three other networks: the international trade network, military dispute network, and the alliance network.



INDIVIDUAL LEVEL


For the primary individual level analysis, two dependent variables are of concern: the total number of recommendations received by a state (in-degree) and the total number of recommendations provided by a state (out-degree). 195 states are participating in the UPR. However, the data consists of 189 states due to missing and lack of data for some states: Andorra, Holy See, Micronesia, Palestine, Serbia, and South Sudan. It should also be noted that for the multilinear regression analysis, the number of observations consists of only 159 states because of missing data for the Polity2 score.3


From Table 3 and Table 4, it can be observed that there is variation in the number of recommendations received and provided. Multiple linear analysis will be conducted to realize which factors influence the variation. From the rankings, it is difficult to discern common state attributes that lead to receiving more or giving more recommendations.


Next, a model focusing on the network property betweenness centrality will be analyzed to understand which states should be considered as the gatekeepers of information or mediators in the UPR process. . While several network properties are measuring the centrality of an actor, such as in-closeness (-0.942), out-closeness (-0.575), in-eigenvector (0.988), and out-eigenvector (0.958), these measures demonstrated a high correlation (correlation coefficients in parentheses) with in-degree and out-degree. Betweenness centrality’s correlation with in-degree (0.409) and out-degree (0.737) is relatively lower than the other properties providing the reason for further analysis.


Betweenness centrality measures the control of information or brokerage role by a node based on the concept of local dependency. If a node B is in the path of A to C, and A must pass by B to get to C and vice versa, then it could be stated that A and C are dependent on point B. Betweenness centrality is calculated halving by the sum of betweenness proportion between a pair of points, known as pair dependency . Betweenness proportion is the proportion of geodesics connecting pair A and C that passes through B (Scott 2013). In other words, betweenness centrality measures the frequency of how often a node is included in the shortest path of one node to another node, or how often it connects nodes that otherwise would have had structural holes.


Due to the relational and dependent nature of betweenness centrality, the multiple linear regression proceeds in two-steps utilizing the UCINET 6 program. First, a standard multiple regression is performed across the dependent (betweenness centrality) and independent vectors. Second, the dependent vector’s values are randomly permuted and recomputed 10,000 times to estimate the standard errors. While the coefficients are calculated through standard OLS linear modeling, standard errors are estimated by stimulation.



EXPLANATORY VARIABLES NETWORK LEVEL


Three networks were constructed from dyadic data on each of the measures. All valued and binary versions of the networks were created for the Quadratic Analysis Procedure (QAP)correlation regression and multiple regression QAP (MR-QAP). The valued ties in the trade network represent the amount of import and export from one state to another state in millions of dollars in the year 2007. For example, the in-degree of country A is the amount of import from country B, and the out-degree of country A is the amount of export to country B. The military dispute5 network is also valued and directed, so country A’s in-degree represents the number of military disputes that the country was a target of, and the out-degree counts the number of initiated military disputes. The military dispute network thus depicts the relationship of conflict representing which states went to war. The alliance network is undirected to discount the information on who initiated the alliance in this research. Ties between two states represent the total number of alliances, including the defense pact, neutrality or non-aggression treaty, or entente agreement. All military disputes and alliances that terminated between the period 1900 to 2010 (military disputes) or 2012 (alliances) are taken into account because the history of enmity and friendship has a lasting effect that defines current relationships. I obtained the data from the Correlates of War (COW) data project including the following datasets: Trade dataset; Dyadic Militarized Inter-State Dispute dataset; and Formal Alliances dataset.



INDIVIDUAL LEVEL


To test the effects of national power, I use the Composite Index of National Capabilities (CINC) score, an index of aggregated six national material capabilities: iron and steel production, military expenditures, military personnel, energy consumption, total population, and urban population. The CINC reflects an average of a state’s share of the world total for every single value per year. Export variable measures a state’s total amount of exports to the world indicating economic power and ties (COW Trade data). Also, a measurement of power is the number of wars initiated by a state; targeted war is the number of wars the state was a target of (COW Militarized Interstate Dispute data). The number of alliances is ally an indicator of military ties and connections (COW Alliances data).


The control variables are indicators of human rights records and compliance to human rights captured with the CIRI Physical Integrity Rights Index and the number of ratified nine core human rights agreements and optional protocols (Fariss 2017). The Physical Integrity Rights Index is constructed from torture, extrajudicial killing, political imprisonment, and disappearance indicators. The Polity2 index represents regime type ranging from -10 (autocracy) to 10 (democracy) (Marshall, Gurr, and Jaggers 2013). According to Hong (2018), treaty compliance is positively associated with providing more meaningful recommendations. US Agreement variable is Lijphart’s index of agreement between the state and the U.S. (UN Ideal Point data).

FINDINGS

COMPARISON OF NETWORKS


In explaining the structure of the UPR network, this section is concerned with the correlation between the UPR network and three other networks: (1) the international trade network, (2) military dispute network, and (3) the military alliance network. The Quadratic Assignment Procedure (QAP)6 is a statistical test that determines the correlation coefficients for two networks. The QAP correlation enables us to perceive the similarities or differences between the UPR network and the trade and alliance networks. If the ties formed in the UPR network are independent of military alliances and trade relationships, the QAP results will need to confirm that the networks have a low correlation.


The results of the QAP correlation are reported in Table 8 with the Pearson correlations and the significance measure in parentheses. I use a bootstrapping test by performing 5,000 permutations through the UCINET 6 program, thereby creating a sample of 5,000 correlation coefficients from a null distribution where there is no association.


The Pearson coefficients measure the similarity between value networks. Although the correlation coefficients are highly significant for the military dispute network and the alliance network, the observed values of the correlation coefficients are low. Commonly, a p-value lower than 0.05 is regarded as statistically significant and marked with an asterisk in the above table (Scott 2013). A low but significant correlation indicates that overlap for ties exists, but the multiplexity of the relations should be more thoroughly included in the analysis (De Lange et al. 2004). As directional networks, recommendations provided and received by states have a more complicated relationship regarding importing and exporting, initiating and participating in military disputes and alliances. Therefore, further analysis in the next section decomposes the network into individual node level features to clarify the correlation among diverse factors to the number of recommendations provided and received. Nonetheless, QAP test results demonstrate that at the structural level, the UPR network has a relatively low correlation with the international trade network, military dispute network, and the alliance network.


Furthermore, due to the low Pearson correlation values, I transformed the value networks into binary networks to analyze the matching coefficients. The Pearson correlation takes into account the total count of ties, whereas the matching coefficient is a reasonable measure for binary networks composed of data on a tie (=1) or no tie (=0). There is an observed simple matching of 0.781 between the UPR network and the military dispute network. This signifies that if there is a relationship between two states in the UPR network, there is a 78% chance that there will also be a relationship between the corresponding states in the military dispute network. However, matching randomly re-arranged matrices reported an average of 0.758, so the observed measure differs by only a small degree from a random result. The average value reports the average correlation of matching random actors across a large number of random permutation trials. The same explanation applies to the trade and alliance matching coefficients, where the observed coefficient is relatively high, but the averaged coefficient is also similarly high.


Additional analyses with the Multiple Regression Quadratic Assignment Procedure (MR-QAP) yielded similar results. MR-QAP is a network regression technique that enables us to know the effect of the independent variables on the dependent network. We can also predict if the strength of ties in trade, military disputes, and alliances increase the likelihood of interaction in UPR. Standardized coefficients for the effect of the trade network (0.021*), military dispute network (0.095***), and alliance network (0.027**) are statistically significant, but the magnitude of the coefficients are small, and R-square for all three models was near zero.


Conclusively, the QAP and MR-QAP analyses demonstrate that the UPR network has a low correlation with the military dispute network and the alliance network. The observed correlation in the data is more significant than 95% of the correlations from permuted random data. Thus, there is a significant association between the UPR network and the military dispute network, and between the UPR network and the alliance network. So UPR interactions could be understood as partially associated with military dispute relations and alliance ties. However, the trade network shows no significant association with the UPR network. That is, imports and exports among nation states’ structural network do not correlate with the interactions in UPR. To further understand the factors influencing state interaction in the UPR, a linear regression test with individual node level network attributes and independent variables measuring state power and human rights indicators are conducted in the next section.



THE RELATIONSHIP AMONG STATE POSITION, POWER, AND COMPLIANCE TO HUMAN RIGHTS


The second analysis focuses on the determinants of UPR participation and recommendations; that is, “what factors explain the variation in in-degree (number of recommendations received by the state) and out-degree (number of recommendations provided by the state)?” While the in-degree is inversely correlated with national power variables following Hypothesis 2, findings are not statistically significant. Instead the Physical Integrity index makes receiving and giving recommendations more likely. The results confirm Hypothesis 3 that states with more national power are more likely to provide recommendations.


For both models, the findings show that the effect of the Physical Integrity index is statistically significant. However, the coefficient directions are opposite as expected. A state with a low Physical Integrity index is more likely to receive more recommendations from other states. A one-unit change in in-degree leads to a decrease of 1.2 in the Physical Integrity Score, that ranges from -8 (no government respect for rights) to 8 (full government respect for rights on torture, political imprisonment, disappearance, extrajudicial killing). On the other hand, a state with a high Physical Integrity index is more likely to provide more recommendations to other states. A one-unit change in out-degree leads to an increase of 4 in the Physical Integrity Score. This finding is logically coherent with our understanding that states with low human rights records should receive more criticism and recommendations, whereas states with high human rights records should provide more recommendations to other states. Contrastingly, the number of ratified HRA has no statistically significant effect on the state’s level of participation.


Because of the directional nature of the UPR data, different variables influence why a state receives or provides recommendations. For the in-degree model, the Polity2 score, Physical Integrity index, and Lijphart’s Index of Agreement’s percent agreement with the US are statistically significant. We find no support for Hypothesis 2 as the CINC index, trade exports, number of allies, and initiated military disputes do not have an observed effect on the count of recommendations received by a state. Autocratic states are more likely to receive more recommendations than democratic states. A state with higher vote agreement with the U.S. in the UN. This suggests that a friend of the U.S. may be subject to more criticism by others, or that U.S. allies provide more complimentary comments to one another influencing the number of recommendations received. The US agreement score is positively correlated with the Polity2 index, national capability index, and Physical Integrity index, but has a positive correlation with in-degree, opposite the negative correlations of the other related factors. Thus, there exists a possibility that comments received by those agreeing highly with U.S. positions in the UN include compliments or general recommendations with low severity. Previous literature find that states tend to make weak recommendations to political allies (Chow 2017; Terman & Voeten 2018).


As for the hypothesis on out-degree, I find that the effect of national power is stronger than the effect of the Physical Integrity index. A one-unit increase in out-degree is associated with 12.418 increase in national capacity and a 3.93 increase in national exports to the world. Noting that the CINC is an aggregate of military capacities, total population, urban population, and consumption of resources, greater military power, higher population, and higher consumption of resources indicates that the country will provide more recommendations to others. The results support the hypothesis that states interlocked in trade relationships and have high levels of exports, and thus more economic power, provide more recommendations. The findings agree with McMahon & Ascherio’s (2012) observation that aid recipients take caution before criticizing their Western donor states. Finally, states with high physical integrity index, and in turn, a high level of human rights protection, tend to give more recommendations. However, the effect of national capacities is more significant and greater on the out-degree.


In sum, the findings present strong support for Hypothesis 3 on the positive effect of national power’s on the reviewer’s number of recommendations provided. The results do not support the association among national power variables and the in-degree. State regime, state human rights records, and agreement to U.S. votes in the UN account for in-degree variations. The Physical Integrity index served as a predicting factor for both in- and out-degree counts. These results suggest that material power and political relations affect how often a state participates in the UPR, but states recommend and receive criticism based on their human rights records.



BETWEENNESS CENTRALITY POSITION


The final step of the analysis is a test of regressing the national capacities variables on the network centrality measure. Betweenness centrality is a measure derived from a state’s position in the UPR network. Observing the betweenness centrality tells us whether the same variables that influence UPR participation, also help explain a state’s position in the network itself. The same variables utilized in Table 9 are employed in the model, except for Initiated MID and Target MID, which have been aggregated as Total MID for the reason that it increases the statistical significance of the model. However, it should be noted that the p-value of the F-test is 0.105, so the model using the same variables as the in-degree and out-degree is not statistically significant.


Nonetheless, scrutinizing the results demonstrate that the Physical Integrity index and Polity2 index affect the betweenness centrality. This rejects Hypothesis 4 that posits national power induces higher betweenness centrality measures. Similar to the result for in-degree, higher Physical Integrity protection and a greater degree of Authoritarianism predicts higher betweenness centrality.


For example, the country with the highest betweenness centrality measure is Cuba; Cuba’s Polity2 score is -7 and its Physical Integrity score is 4. Betweenness centrality measures the structural positions of brokers within the network, in this case representing the value of participating actively as both a recommender and receiver of human rights comments. Cuba’s high betweenness centrality could be traced back to its spectrum of friends and enemies from a broad political group; that is, Cuba gave comments and received comments from both democratic and authoritarian states as well as those from the North and the South.


Cuba appeals to decolonized and developing states in order to evade criticism (Freedman 2015, 75). Cuba utilizes human rights language to voice its anti-West point of view, which many states sympathize with (Freedman 2015, 75). It also has allies in Eastern Europe due to similar political outlooks. Political or regional allies such as Venezuela and Syria, endorse Cuba’s statements and similarly strategically attack the U.S. (Freedman 2015, 76). In other words, Cuba utilizes its position to avoid criticism and rather to condemn human rights records of other countries.


Other states with the highest betweenness centrality measures are the United States, Austria, Slovenia, Mexico, Malaysia, Turkey, Nigeria, and Australia. The United States, Austria, Slovenia, Mexico, Turkey, and Australia have high Polity2 scores that mark them as democratic states. Contrastingly, Malaysia, Nigeria, Chad, Iran, Zimbabwe have very low Polity2 scores, but high betweenness centralities. In all, the betweenness centrality measures the state with the most control of information that passes through other states. In the UPR network, this means having interacted with states with diverse attributes since the broker states connect the unconnected states.


Cuba and the U.S. have discrepant national capacities and state relations. Nonetheless, the two states have similarly high betweenness centralities. This can be explained in terms of social power as access itself, and the fungibility of network positions. Since the UPR ensures equal participation, states without material capacity may utilize the rules and procedures to garner social power. For example, states with problematic human rights records can request to their allies to sign up for the speaker’s list, which is a first come, first served rule, to avoid time for critical recommendations. During Cuba’s first cycle, only eight out of fifty-three countries asked critical questions on freedom of expression (Kälin 2014, 32). In such ways, Cuba and other states could gain social power as access.


Could this positional power translate into other forms of power? Another aspect of the difference between Cuba and U.S. stems from the fungibility of network power. The U.S. has more freedom than Cuba in terms of deciding upon its human rights foreign policy despite the criticism and attention it receives from other states. Other forms of material power enable the U.S. from loosening the constraints of positional power in the UPR. Freedman (2015) outlines in detail the various ways in which the U.S. justifies its behavior that does not accord to international standards. The U.S. insists that different laws should apply during the War on Terror; it refuses to be bound by international law on the environment, human rights, criminal law, trade agreements and so on (Freedman 2015, 3; 6). However, Cuba is subject to more coercion than the U.S. to abide by international human rights norms.


Nonetheless, for insincere compliers like Cuba, human rights mechanisms are used to supplement their reputational power, gained by joining and participating in the process, but the constraints are minimal leaving small incentives to abide by the recommendations. The fungibility perspective may enable a clearer understanding of why authoritarian states have high betweenness centrality scores in the UPR and extend to explanations on why they participate in international organizations at large. This leaves us with the normative question on the meaning of position and power in the UPR.


In sum, the UPR process needs states with high betweenness centrality to engage with various states in the discourse. If states only spoke selectively to friends or enemies, then the UPR would not be able to carry out its universal, peer-review mechanism properly. Cowan & Billaud (2015) determined the good guys in the UPR as states who do not have critical human rights violations and could assume leadership positions in the public discussions. More so than leader and laggards, the performance of the mediators and gatekeepers may be a more critical role for states to focus on in the future.

CONCLUSION

This study contributes to the literature on the UPR and human rights social network analysis (SNA) in three ways. Primarily, I advance the application of SNA on state actors’ human rights networks . SNA in human rights literature focuses on transnational advocacy networks (TAN) of non-state actors or international nongovernmental organizations (Keck & Sikkink 1998; Murdie & Davis 2012a; Murdie & Davis 2012b). I contend that the network of state actors convey the political relationships among states that constrain and influence state behavior in a multilateral mechanism. Comparing the structural relationships within the UPR network with three networks that depict the flow of economic and military ties, this study finds that the UPR network is mostly independent of those networks on the material, national power.


Second, in conjunction with the network analysis, I evaluate the relationship between the level of state participation, national power, and human rights records. I demonstrate the importance of state human rights records in both behaviors of receiving and providing recommendations. States with higher national capabilities and greater trade exports are more likely to provide recommendations. Nonetheless, while smaller in magnitude, states’ human rights record influences the number of recommendations. Furthermore, states with low human rights records and a lower level of democracy are more autocratic, are more likely to receive recommendations. The influence of Physical Integrity protection scores on both the in- and out-degrees mean that interactions in UPR are driven by human rights logic. Thus, the results imply a conditional politicization where state activity of providing recommendation tends to be explained more by state power and capacities and states receiving recommendations are targeted for low levels of human rights. Both are significantly affected by a state’s human rights performance.


Finally, the network centrality measure, betweenness centrality, is highly correlated with higher levels of human rights records and lower levels of democracy. The result demonstrates that states with high physical integrity scores have a more central position in the UPR interactions. This signifies that the role of gatekeeping is conferred to states with better human rights records to provide recommendations and interact with diverse states actively. On the other hand, states with low levels of democracy also seem to have high centrality scores implying that by receiving recommendations from various states, they received central attention from the UPR process. From this analysis, we can only carefully infer the meaning of the independent variables since the betweenness centrality measure does not take into consideration the direction of the recommendation. Nonetheless, the progress made in this article is that the state’s positions within the network have been measured and understood albeit in a limited vein.


The findings have a significant implication on the politicization of the UPR. Previous literature’s concern on the UPR’s politicization focused on a state’s recommendation giving behavior and the lack of reciprocity within the mechanism. However, network comparisons and regression models depict a smaller influence of trade and political ties at the structural level, and a greater effect of human rights records that subject states to more criticism. This study confirms the presence of politicization, but also ensures that the UPR is functioning according to its main purpose in criticizing the human rights violations of non-compliers. The new revelation is that the UPR network contains different ties with trade and military relationships and that the role of countries with high betweenness centrality should receive more scholarly attention to improve the peer-review function.

Tables
Table. 1. Univariate Descriptive Statistics for First and Second Cycle UPR
Table. 2. Network level measures for First Cycle UPR Directed/Valued Network
Table. 3. Individual level Dependent Variable Summary
Table. 4. In-degree and Out-degree Rank4
Table. 5. Node level Network Property Dependent Variable Summary
Table. 6. Univariate Statistics for Three Networks
Table. 7. Individual level Independent and Control Variable Summary
Table. 8. QAP-Correlation Matrix
Table. 9. Multiple Linear Regression Model for In-degree and Out-degree
Table. 10. Multiple Linear Regression Model for Betweenness Centrality
Footnotes
1)In scrutinizing the politicization of the World Trade Organization (WTO), Reich (2005) defines politicization as “a situation where actions are taken for purposes unrelated or inadequately related to the goals and functions of that IGO, but rather stem from the geopolitical goals and strategies of a particular member state or group of member states” (784). Ghebali (1985) identifies six types of dysfunctions leading to politicization in the UN specialized agencies: (1) extraneity, the systematic insertion of extraneous issues into the debate; (2) hyper-confrontation, changing the rules of debate; (3) excommunication, the near-expulsion of member states by singling out certain states by the majority with a double standard; (4) erosion of liberal principles included in the IGO charters; (5) nomomania, the practice of forcing resolution of one-sided normative content to pass; (6) administrative mismanagement and budgetary excesses. Lyons et al. (1977) provide conceptual ideas of politicization as (1) a defect to be corrected, an organization defect; (2) an indicator of environmental forces bearing on the organization;(3) a bargaining process that prioritizes the needs of, for example, third world nations (88).
2)Lyons et al. (1977) quote Charles William Maynes to incorporate the idea of the North-South confrontation to politicization in specialized agencies.
3)Nonetheless, the model has been tested without the Polity2 score with 189 states and the coefficient magnitude and direction showed little variation. It is safe to assume that the data size of 159 states shows similar results with the sample of 189 states.
4)It should be noted that the network does not include six countries. Therefore, recommendations provided by the excluded six countries and the recommendations given to the excluded six countries are not included in the dataset. This affects the total number of recommendations received and provided of the other 189 countries included in the research dataset if they had received or given the recommendation to the excluded six countries.
5)Only disputes that meet one of the conditions of 100 battle related fatalities and 1,000 or more troops deployed by both states are considered a dyadic military dispute.
6)The correlation coefficients for one of the networks and a randomly re-arranged version of the other network permuted thousands of times are calculated to determine whether the randomized network differs from the observed network (Scott 2013, 144).
References
  1. Bailey, Michael A., Anton Strezhnev, and Erik Voeten. 2017. “Estimating Dynamic State Preferences from UN Voting Data.” Journal of Conflict Resolution 61(2), 430-456.
  2. Barbieri, Katherine, Omar M. G. Keshk, and Brian M. Pollins. 2009. “Trading Data: Evaluating our Assumptions and Coding Rules.” Conflict Management and Peace Science 26(5), 471-491.
  3. Barbieri, Katherine and Omar M. G. Keshk. 2016. Correlates of War Project Trade Data Set Codebook, Version 4.0 . Accessed at http:// correlatesofwar.org (August 20, 2018).
  4. Boehmer, Charles, Erik Gartzke, and Timothy Nordstrom. 2004. “Do Intergovernmental Organizations Promote Peace?” World Politics 57(1), 1-38.
  5. Brown, Bartram S. 2011. The United States and the Politicization of the World Bank: Issues of International Law and Policy . New York: Routledge.
  6. Carraro, Valentina. 2017. “The United Nations Treaty Bodies and Universal Periodic Review: Advancing Human Rights by Preventing Politicization?” Human Rights Quarterly 39(4), 943-970.
  7. Chow, Jonathan T. 2017. “North Korea’s Participation in the Universal Periodic Review of Human Rights.” Australian Journal of International Affairs 71(2), 146-163.
  8. Cingranelli, David L. and David L. Richards. 1999. “Measuring the Level, Pattern, and Sequence of Government Respect for Physical Integrity Rights.” International Studies Quarterly 43(2), 407-18.
  9. Cingranelli, David L., David L. Richards, and Chad K. Clay. 2014. The CIRI Human Rights Dataset . Accessed at http://www.humanrightsdata.com (August 20, 2018).
  10. Cochrane, Leanne and Kathryn McNeilly. 2013. “The United Kingdom, the United Nations Human Rights Council and the first cycle of the Universal Periodic Review.” The International Journal of Human Rights 17(1), 152-177.
  11. Cowan, Jane K. and Julie Billaud. 2015. “Between Learning and Schooling: The Politics of Human Rights Monitoring at the Universal Periodic Review.” Third World Quarterly 36 (6), 1175-1190.
  12. De Lange, Daniëlle, Filip Agneessens and Hans Waege. 2004. “Asking Social Network Questions: A Quality Assessment of Different Measures.” Metodološki Zvezki 1(2): 351-378.
  13. Dominguez-Redondo, Elvira. 2012. “The Universal Periodic Review - Is there Life beyond Naming and Shaming in Human Rights Implementation.” New Zealand Law Review 4, 673-706.
  14. Dorussen, Han and Hugh Ward. 2008. “Intergovernmental Organizations and the Kantian Peace: A Network Perspective.” Journal of Conflict Resolution 52(2), 189-212.
  15. Dorussen, Han and Hugh Ward. 2010. “Trade networks and Kantian peace.” Journal of Peace Research 47(1), 29-42.
  16. Fariss, Christopher J. 2017. “The Changing Standard of Accountability and the Positive Relationship between Human Rights Treaty Ratification and Compliance.” British Journal of Political Science 48(1), 239-271.
  17. Finnemore, Martha and Kathryn Sikkink. 2001. “Taking Stock: The Constructivist Research Program in International Relations and Comparative Politics.” Annual Review of Political Science 4(1), 391-416.
  18. Freedman, Rosa. 2013. The United Nations Human Rights Council: A Critique and Early Assessment . New York: Routledge.
  19. Freedman, Rosa. 2015. Failing to Protect: The UN and the Politicisation of Human Rights . Oxford: Oxford Univ. Press.
  20. Freedman, Rosa and Ruth Houghton. 2017. “Two Steps Forward, One Step Back: Politicisation of the Human Rights Council.” Human Rights Law Review 17, 753-769.
  21. Gaer, Felice D. 2007. “A Voice Not an Echo: Universal Periodic Review and the UN Treaty Body System.” Human Rights Law Review 7(1), 109-139.
  22. Gartzke, Erik and Kristian S. Gleditsch. 2008. “The ties that bias: Specifying and operationalizing components of dyadic dependence in international conflict.” Manuscript, University of California, San Diego. Accessed at http://pages.ucsd.edu/~egartzke/papers/tiesthatbias_05192008.pdf.
  23. Ghebali, Victor-Yves. 1985. “The Politicisation of UN Specialised Agencies: A Preliminary Analysis.” Millenium: Journal of International Studies 14(3), 321-325.
  24. Gibler, Douglas M. 2009. International military alliances, 1648-2008 . Washington D.C.: CQ Press.
  25. Hafner-Burton, Emilie M. and Alexander H. Montgomery. 2006. “Power Positions: International Organizations, Social Networks, and Conflict.” The Journal of Conflict Resolution 50(1), 3-27.
  26. Hafner-Burton, Emilie M. and Alexander H. Montgomery. 2012. “War, Trade, and Distrust: Why Trade Agreements Don’t Always Keep the Peace.” Conflict Management and Peace Science 29(3), 257-278.
  27. Hafner-Burton, Emilie M., Miles Kahler, and Alexander H. Montgomery. 2009. “Network Analysis for International Relations.” International Organization 63(3), 559-592.
  28. Haim, Dotan A. 2016. “Alliance networks and trade: The effect of indirect political alliances on bilateral trade flows.” Journal of Peace Research 53(3), 472-490.
  29. Hegre, Havard, John R. Oneal, and Bruce Russett. 2010. “Trade does promote peace: New simultaneous estimates of the reciprocal effects of trade and conflict.” Journal of Peace Research 47(6), 763-774.
  30. Hong, Mi Hwa. 2018. “Legal Commitments to United Nations human rights treaties and higher monitoring standards in the Universal Periodic Review.” Journal of Human Rights , 1-14.
    CrossRef
  31. Ingram, Paul, Jeffrey Robinson, and Marc L. Busch. 2005. “The Intergovernmental Network of World Trade: IGO Connectedness, Governance, and Embeddedness.” American Journal of Sociology 111(3), 824-858.
  32. Jackson, Matthew O. and Stephen Nei. 2015. “Networks of Military Alliances, Wars, and International Trade.” Proceedings of the National Academy of Sciences , 1-8.
  33. Kälin, Walter. 2014. “Ritual and Ritualism at the Universal Periodic Review: A Preliminary Appraisal.” In Hilary Charlesworth and Emma Larking eds., Human Rights and The Universal Periodic Review: Rituals and Ritualism, Cambridge: Cambridge University Press.
  34. Landolt, Laura K. 2013. “Externalizing Human Rights: From Commission to Council, the Universal Periodic Review and Egypt.” Human Rights Review 14, 107-129.
  35. Lebovic, James H. and Erik Voeten. 2006. “The Politics of Shame: The Condemnation of Country Human Rights Practice in the UNCHR.” International Studies Quarterly 50(4), 861-888.
  36. Lupu, Yonatan and Vincent A. Traag. 2012. “Trading Communities, the Networked Structure of International Relations, and the Kantian Peace.” Journal of Conflict Resolution 57(6), 1011-1042.
  37. Lyons, Gene M., David A. Baldwin, and Donald W. McNemar. 1977. “The ‘Politicization’ Issue in the UN Specialized Agencies.” Proceedings of the Academy of Political Science 32(4), 81-92.
  38. Maoz, Zeev, Paul L. Johnson, Jasper Kaplan, Fiona Ogunkoya, and Aaron P. Shreve. 2019. “The Dyadic Militarized Interstate Disputes (MIDs) Dataset Version 3.0: Logic, Characteristics, and Comparisons to Alternative Datasets.” Journal of Conflict Resolution (forthcoming).
  39. Maynes, Charles William. 1976. “A U.N. Policy for the Next Administration.” Foreign Affairs 54, 804-819.
  40. McMahon, Edward R. and Marta Ascherio. 2012. “A Step Ahead in Promoting Human Rights? The Universal Periodic Review of the UN Human Rights Council.” Global Governance: A Review of Multilateralism and International Organizations 18(2), 231-248.
  41. Milewicz, Karolina M. and Robert E. Goodin. 2016. “Deliberative Capacity Building through International Organizations: The Case of the Universal Periodic Review of Human Rights.” British Journal of Political Science 48(2), 513-533.
  42. Murdie, Amanda and David R. Davis. 2012a. “Looking in the mirror: Comparing INGO networks across issue areas.” The Review of International Organizations 7, 177-202.
  43. Murdie, Amanda and David R. Davis. 2012b. “Shaming and Blaming: Using Events Data to Assess the impact of Human Rights INGOs.” International Studies Quarterly 56(1), 1-16.
  44. Poast, Paul. 2010. (Mis)Using dyadic data to analyze multilateral events. Political Analysis 18(4), 403-425.
  45. Reich, Arie. 2005. “The Threat of Politicization of the WTO.” University of Pennsylvania Journal of International Economic Law 26(4), 779-814.
  46. Russett, Bruce and John R. Oneal. 2001. Triangulating Peace: Democracy, Interdependence, and International Organizations . New York: Norton.
  47. Schaefer, Brett and Steven Groves. 2010. “The U.S. Universal Periodic Review: Flawed from the Start.” The Heritage Foundation . Accessed at https://www.heritage.org/report/the-us-universal-periodic-review-flawed-thestart (August 23, 2018).
  48. Scott, John. 2013. Social Network Analysis . Los Angeles: Sage.
  49. Smith, Rhona K. M. 2011. “Equality of ‘Nations Large and Small’: Testing the Theory of the Universal Periodic Review in the Asia-Pacific.” Asia- Pacific Journal on Human Rights and the Law 2, 36-54.
  50. Soh, Changrok and Kim Yooneui. 2014. “UN Charter-Based Human Rights System and East Asian Human Rights - Focusing on the Universal Periodic Review.” The Korean Journal of International Studies 54(4), 159-193.
  51. Terman, Rochelle and Erik Voeten. 2018. “The Relational Politics of Shame: Evidence from the Universal Periodic Review.” Review of International Organizations 13, 1-23.
  52. UPR-Info. 2018. Database of UPR Recommendations. Accessed at https://www.upr-info.org (July 4, 2018).


17-2 (August 2019)