Legislative Process in International Organizations

Abstract

In this analytical essay, we advance a simple but powerful claim: scholars can better understand outcomes of international organizations (IOs) by developing theories that explicitly make assumptions about legislative process. Because process assumptions powerfully explain domestic legislative outcomes and many international assemblies demonstrate similarities to domestic legislatures, scholars could usefully employ legislative-process-centric approaches when theorizing about outcomes in world politics. Following an explication of why scholars might focus on legislative procedure, we describe several legislative procedures and highlight variance across those procedures within several well-known IOs. We also suggest that this variance and the shadow of power politics cast over IOs provides fertile ground for comparative legislative scholars—including scholars of the U.S. Congress—to develop and test new theories of legislative procedure.

In this analytical essay, we advance a relatively simple but powerful claim: scholars can better understand outcomes in international organizations (IOs) by developing theories that explicitly make assumptions about legislative process. Because (1) we know from research on domestic legislatures that process assumptions powerfully explain legislative outcomes (e.g., Shepsle and Weingast 1994), and (2) many international assemblies are similar in terms of process to domestic legislatures,¹ international relations scholars could usefully adopt legislative-process-centric approaches to developing theories about myriad outcomes in world politics.

In recent years, scholars have focused much attention on the outcomes produced by international assemblies—for example, country-vote choice in the United Nations General Assembly (UNGA) plenary (e.g., Voeten 2013) or the number of loans distributed by the World Bank (e.g., Dreher, Sturm and Vreeland 2009).²Martin and Simmons (1998) suggested the focus on the outcomes of IOs to be a consequence of international relations scholars borrowing from literature on the U.S. Congress in the 1980s:

Perhaps due to new and exciting work in U.S. legislative behavior, the IO research program quickly became focused on how to describe patterns of voting in the (UN) General Assembly, without a systematic attempt to sort out the usefulness of the voting behavior approach. Despite warnings that the international system was fundamentally different from domestic political systems, this research program easily accepted that voting in the UN was a proxy for power in that institution.

Domestic legislatures and international assemblies demonstrate some fundamental differences, making it perilous to examine outcomes or procedures in IOs as though they are simply domestic legislatures. Yet we argue that a fundamental insight from domestic theories—that process is a useful framework for building theories of legislative outcomes—could be very fruitful for scholars interested in explaining behavior in IOs. Following Martin and Simmons’ (1998) call for a “systematic attempt to sort out the usefulness of the voting behavior approach,” Milner’s (1998) insistence that “to understand the major issues in international politics...IR theorists must bring a systematic analysis of domestic politics into the field,” and Hafner-Burton, von Stein, and Gartzke’s (2008) reminder of the importance of “shifting the focus from whether IOs matter to how they work,” we suggest that international relations scholars can most benefit from the insights of the literature on domestic legislatures by focusing on two key areas.

First, perhaps the most fundamental lesson of scholarship on domestic legislatures is that process is a key tool by which power is translated into influence (e.g., McKelvey 1976; Shepsle and Weingast 1994). By process, we mean the institutional rules—like agenda control, committee consideration, amendments, and roll call votes—by which a proposal becomes a formalized and finalized action of an assembly. Indeed, the concept of structure-induced equilibrium suggests that we can only link voting outcomes to individual preferences through an understanding of process (Shepsle and Weingast 1981). Recent scholarship on outcomes in European Union (EU) provides a useful template for the consideration of legislative process; Tsebelis and Garrett (1996, 1997), for example, describe EU outcomes as a function of voting rules and the extent to which actors wield agenda-setting and veto power.³

In this essay, we focus on process, but we recognize that process is subsequent to—and arguably synonymous with—institutional design. Indeed, legislative process is simply the downstream effect of legislative organization.⁴ We do not focus on the origination of institutions—or the extent to which actors can strategically design institutions and processes to get what they want. Others have carefully considered this endogeneity, both in the international (e.g., Koremenos, Lipson and Snidal 2001a,b; Rosendorff and Milner 2001; Stone 2011) and domestic contexts (e.g., Riker 1980; Sinclair 2016).

Second, scholars interested in a particular explanatory concept might gain extra leverage by treating legislative process as an intervening factor in their theories. For example, Thompson (2006) incorporates Krehbiel’s (1991) assumptions about the effects of different types of legislative committees into his theory about the informational advantages of channeling coercive policies through IOs. But a wide variety of literature looking at the effect of some factor, x, on some outcome, y, fails to consider the extent to which such relationships are mediated by legislative process. For example, measures of state foreign policy preferences created using data on UNGA votes (e.g., Voeten 2000) have been used to predict a wide variety of international outcomes. But by relying only on votes—rather than also taking into consideration the legislative process by which those votes were created—scholars may be mischaracterizing the revealed preferences of states.

We recognize that scholars are aware of the importance of process in the study of IOs; references to the literature on domestic legislatures are relatively commonplace in the IO literature. Yet there are relatively few examples of theories that rigorously incorporate assumptions about legislative process to explain outcomes in IOs. To illustrate our point, consider that few analyses of voting outcomes in the United Nations Security Council (UNSC) ignore the fact that five powerful countries wield veto power; this process structures the extent to which countries are able to influence Security Council outcomes. Russett, O’Neill and Sutterlin (1996) provide a thorough analysis of how UNSC decision-making might change as a result of proposed restructuring of the veto. Because of the power of the veto, UNSC outcomes often differ dramatically from those in the UNGA—even on the same substantive issues.

While our example regarding the effects of UNSC veto power is obvious to the point of being rudimentary, few analyses consider other aspects of the UNSC’s legislative process, either as a main or a modifying variable.⁵ Legislative process in international assemblies beyond the UNSC and UNGA also varies dramatically. Even with substantial variance to explain, few theories explicitly and centrally make assumptions about how procedures—how committees are organized, how votes are aggregated, who has the power to set the agenda, and which procedures govern meetings and deliberation—affect outcomes and which countries have the power to influence those outcomes. While we hesitate to argue that these elements of process influence outcomes as starkly as the P5’s veto, it is difficult to imagine that procedures so strongly affecting outcomes in domestic legislatures would have no effect in international assemblies.

Legislative Processes and Variation in IOs

In the remainder of this letter, we consider several aspects of process argued to be important in domestic legislatures: committees, voting rules, agenda setting, and meeting powers. For each procedure, we organize our discussion around three questions: (1) What is the procedure?; (2) Why is it important?; and (3) Which IOs have it? Table 1 provides a summary of the variance in these procedures across our convenience sample of international assemblies.⁶ Additional information about our coding procedures and decisions is available in our Supplemental Appendix.

Committee Membership

Committees are internal legislative bodies—typically composed of subsets of members of a full legislature—to which action on particular topics is delegated. The workload of the legislative body is often delegated to individual committees and subcommittees who deal with specialized issues with fixed jurisdictions (e.g., Polsby 1968; Gilligan and Krehbiel 1990; Krehbiel 1991; Squire 1992). Committees promote a division of labor, encourage the development of policy specialization, and increase the efficiency of law making. Because much of the work of drafting and molding proposals occurs in committee, they also serve as key agenda control institutions (e.g., Fenno 1973; Shepsle and Weingast 1987; Gilligan and Krehbiel 1990; Bowler and Farrell 1995; Hage 2007); many proposals see their demise while indefinitely awaiting committee action (e.g., Cox and McCubbins 2005; Den Hartog and Monroe 2019).

Committees play two key roles in theories of legislative power and outcomes. First, committees are information producers (Krehbiel 1991; MacNeil 1963); committee members become experts in a particular area and thus act as informants for legislators who are part of the larger legislative body but not on a particular committee (Shepsle and Weingast 1987; Bianco 1997). Second, committees act as gate keepers, deciding which proposals—and which versions of those proposals—make it to the next stage of process for consideration by a full legislature (e.g., Shepsle and Weingast 1987; Aldrich 1994; Pearson and Schickler 2009).

In the second column of Table 1, we examine variation in the structure of committees in international assemblies. Committees is coded “Whole” if all committees in a given IO are committees of the whole (i.e., the full membership of the body is represented on every committee) and “Subset” if committees are made up of a subset of the body’s whole membership. We would have coded IOs with no committees, but in our sample, every single international legislature operates with committees. In international assemblies where committee membership is limited to a subset of nation-states, IOs scholars might be interested in the extent to which membership on some committees and not others increases a country’s power over particular policy jurisdictions.

In the literature on domestic legislatures, committees of the whole are rare. Consequently, international assemblies provide interesting variance in committee structure that is unavailable in most domestic legislative settings. In the domestic context, where legislatures have standing committees with strong substantive property rights, final adoption votes tend to be closer to unanimous as legislators logroll to see their own committee’s proposals on the floor (see, for example, Shepsle and Weingast 1994). One might expect less unanimity on floor votes in international bodies where committees function as committees of the whole.

Voting

There is enormous variation across legislatures around the world with respect to the means by which votes are cast, the weights attached to those votes, and the method by which those votes are aggregated or translated into decisions. In what follows, we discuss two aspects of voting process relevant to the study of international assemblies: Recorded Votes and Voting Rules.

Recorded Votes

Most legislative bodies have processes to allow for both recorded and unrecorded votes. Informal vote casting includes decisions by consensus or unanimity in which a vote is not recorded. Formal vote casting requires a legislator to affirmatively state his/her support or opposition to a proposal; that statement then results in a recorded vote. Recorded votes are important because they generate a public record of legislators’ positions on policy proposals and increase legislative accountability and responsiveness to the public (e.g., Carey 2003; Carrubba et al. 2006; Poole and Rosenthal 2007; Levy 2007; Snyder and Ting 2008; Hug 2010; Crisp and Driscoll 2012). In addition, recorded votes allow legislators to send signals to interested constituencies including parties and relevant interest groups (e.g., Mayhew 1974; Groseclose, Levitt and Snyder 1999).

In the third column of Table 1, we code Recorded Votes as “Some” if an international legislature records some votes (i.e., some votes are made informally) and “None” if an international legislature records no votes (i.e., all votes are made informally). In our sample, the majority of institutions do not record votes. Importantly, in some of the organizations coded as “None” in Table 1, all voting decisions are made by unanimous consent. Not only is this interesting cross-sectional variation, but previous comparative scholarship has demonstrated (lack of) recorded votes may be used by legislators as a strategic tool (e.g., Arnold 1990; Carrubba et al. 2006; Hug 2010). We might expect to see recorded votes occur more often in international assemblies where signaling to a domestic audience is a crucial element of international participation. Moreover, we might expect to see countries with interested domestic audiences make more requests for recorded votes. Our sample includes six assemblies that make use of recorded and unrecorded votes under different circumstances, providing scholars fertile ground for extending and testing theories regarding the strategic use of roll call votes.

Threshold

Once votes are cast, there is variation in the rules regarding the number of votes needed for a particular motion to carry. In order to pass proposals, many legislatures require a simple majority of votes. Others require a super-majority, commonly two-thirds of the voting members or all of the voting members (i.e., unanimity).⁷ Voting rules are hugely consequential. There is a voluminous literature in studies of legislative (and non-legislative) simple majority voting rules that focuses on the median tendency produced by a simple majority vote (e.g., Downs 1957; Krehbiel 1991; Groseclose and Snyder 1996; McGann 2004; Vermeule 2007; Fish and Kroenig 2011). Conversely, the presence of a supermajority rule mitigates the median tendency and has a tendency to produce gridlock (Krehbiel 1998).

The fourth column of Table 1 shows considerable variation in voting rules in international assemblies. We code Threshold as “Majority” if an assembly uses a majority voting rule, as “2/3” if an assembly uses a two-thirds voting rule, as 2 if an assembly uses a unanimity voting rule, and as “Unanimity” if an assembly uses some other voting rule. Although the majority of IOs in our sample use a majority voting rule, two IOs use an unanimity rule. Other organizations have what we would consider unconventional voting rules given our understanding of domestic legislatures. For example, the UNSC requires the approval of four of the ten temporary members and all five permanent members. To our knowledge, no domestic legislature has voting rules of this sort; as a result, international assemblies provide an opportunity for ground-breaking theoretical and empirical work on the effect of voting rules on strategic behavior and legislative outcomes. Empirically, the larger the voting threshold, the less sensitive outcomes are expected to be in response to shifts in voting membership (Krehbiel 1998). Moreover, as our UNSC example at the outset of this letter suggests, voting rules significantly influence outcomes in international assemblies; IO scholars might wish to account for influence of voting rules—and the potential for gridlock (e.g., Russett, O’Neill and Sutterlin 1996; Binder 2004)—as they develop theories of international influence.

Agenda Setting

The power to determine which proposals make it to a final vote and what form they take when they get to a final vote—in other words, the power to set the agenda—is among the most important factors in determining the distribution of power in a legislative body. In what follows, we discuss two aspects of process related to agenda setting: Agenda Control and Amendments.

Agenda Control

By agenda control, we refer to the ability of an actor, typically a member of the majority or a controlling coalition within a legislature, to influence the process by which a proposal is considered. For example, many theories of legislative decision-making include the concept of “negative agenda control,” which captures the various gate-keeping institutions that allow agenda setters to push some proposals forward while blocking others from consideration. Similar to a super-majority voting rule, negative agenda control can generate gridlock; that gridlock is biased toward the protection of status quos preferred by the agenda setter (e.g., Farquharson 1969; Talbert and Potoski 2002; Cox and McCubbins 2005; Anzia and Jackman 2013). Alternatively, positive agenda control creates movement of status quos toward the preferred position of the agenda setter (e.g., Romer and Rosenthal 1978).

In the Agenda Control column of Table 1, we code agenda control as “Leaders” if only a small subset of leaders has exclusive power to propose agenda items in a given international legislature, as “All” if all members have an equal ability to propose items and have them considered, and as “Coalition” if a coalition of members is required to propose agenda items. Although it is common in many domestic legislatures for leadership to have complete control over the agenda, there are few international assemblies where that is the case. Several of the international assemblies in our sample require a coalition of members to propose agenda items; often where the size of the coalition needed to propose agenda items matches the size of the coalition required to pass a vote, tendencies like the median voter theorem will also govern the effects of agenda power (Krehbiel 1991,, 1998).⁸ When scholars fail to account for agenda setting in theories of legislative decision making, the bulk of the power distribution in that chamber is hidden from the eye of the researcher. Thus, perhaps more than any other aspect of legislative process, IO scholars could expect to make great gains in their understanding of the translation of power to influence by accounting for agenda control.

Amendment Power

By amendment power, we refer to the ability of a member to suggest changes to proposals at some point (typically in committee or on the floor) during the legislative process. While the content of an initial legislative proposal is important, much or all of the substance of a proposal can be changed by an amendment; the amended version can oftentimes look very different than the initial proposal. Thus, the ability to make—or restrict—an amendment can heavily determine which members enjoy the lion’s share of the benefit (e.g., Wilkerson 1999; Marshall 2019; Finocchiaro and Jenkins 2008; Penn 2008; Lynch, Madonna and Roberts 2016). Empirically, following Magleby, Monroe and Robinson (2018), we expect international assemblies that allow amendments to see heightened adoption of minority-coalition proposals. We should also observe larger final adoption coalitions in international bodies that feature amendment rights.

In Column 6 of Table 1, we code Amendments as “All” if all members of a given international legislature can propose amendments to proposals and “Other” otherwise. Although there is little variation on amendment proposal power depicted in Table 1, we include the legislative procedure in our assessment because of both the importance of amendment power in the literature on comparative legislatures and because there is variation with regard to the rules governing when and how amendments can be proposed. In the UNGA, for example, amendments to votes on the plenary floor must be submitted in writing to the UN Secretary General and then passed with a two-thirds vote in the plenary before voting on the proposal [Rule 78]. In the Organization of American States, amendments can be proposed during consideration of the proposal on the floor of the legislature; at that point, votes are cast on both the proposal and the corresponding amendment following the close of discussion [Article 70]. Given that there is interesting variation in the rules governing amendments in the assemblies in our sample, understanding which states have the ability to change proposals strikes us as potentially important in constructing explanations of state power in international assemblies.

Meeting Powers

Legislative business culminates in formal meetings where proposals are considered (i.e., amendments are made, speeches are given, and votes are taken). The ability to determine aspects of these meetings plays a crucial role in determining legislative outcomes.⁹ In what follows, we discuss two aspects of process related to meeting powers: Scheduling Proposals and Debate Limits.

Scheduling Proposals

The power to schedule proposals refers to the prerogative of a member (often a presiding officer or a leader) to determine the order in which proposals are considered in a chamber or committee. Scheduling power can allow the scheduler to realize a bargaining advantage if he/she can prioritize some items and de-prioritize other as a negotiating tactic because it gives an actor ultimatum bargaining power to advance their interest and block policies that are counter to their interest (e.g., Baron and Ferejohn 1989; Ainsworth and Flathman 1995; Yackee 2003; Den Hartog and Monroe 2011). Empirically, we might expect more powerful states to see their proposals considered earlier in meeting cycles in IOs in which they have disproportionate scheduling power.

In Table 1, we code Schedule Proposals as “All” if all members of an international legislature have an equal right schedule a proposal to be considered on the floor, as “Leaders” if only members in leadership positions can schedule a proposal for floor consideration the floor, and as “Unknown” if we were unable to find information on scheduling in the IO’s originating documents. We believe that the large number of international legislatures in our sample that do not mention formal rules for scheduling proposals is due to the use of assembly norms regarding scheduling. For the international assemblies that do have formal rules regarding scheduling, there is important variation; in particular, IO scholars seeking to incorporate scheduling power into their theories should pay careful attention to the strategic ordering of proposals near the end of legislative sessions (Wawro and Schickler 2004).

Debate Limits

Debate limits refer to the ability to restrict the time members are allowed to speak on matters before the chamber and the content (i.e., germaneness) of that speech. Legislators can limit debate on a proposal either unilaterally or through some voting rule (i.e., cloture to end a filibuster in the US Senate). Debate limits are important because they can limit attempts to obstruct proposals and allow for the efficient management of scarce legislative time (Wawro and Schickler 2004; Koger 2006; Oleszek 2014).

In the last column of Table 1, we code Debate Limits as 1 if an international assembly’s formal rules allow for time and/or content limits and as 0 otherwise. Table 1 shows interesting variation with regard to this element of process. In addition, several international assemblies limit debate based on chamber norms rather than formal rules. For example, in the World Trade Organization, we expect that “customary practices” include norms on debate. Stone (2011) provides us with a broad and rigorous framework the “parallel” importance of formal and informal institutions, focusing on the latter and reminding us that actors create institutions strategically: “In order to assure the participation of the most powerful states, international institutions have developed informal procedures that accommodate their interests” ( Stone 2011, 1). This strikes us an area ripe for scholars with in depth case knowledge to contribute to our understanding of how norms and rules shape outcomes in those bodies.

More systematically, as political science moves toward more sophisticated methods for analyzing text, data on speech in IOs will beg for theories informed by scholarship on the limits of legislative debate. Consider the approach to identifying political conflict through texts of legislative speech pioneered by Monroe, Colaresi and Quinn (2008). The arguments in this essay imply that as formal and informal debate limits vary—both over time within a single institution and across different international institutions—we should see change in both the dimensions of verbal disagreement and the associations between states and types of conflict.

Conclusion

Following the lead of Martin and Simmons (1998); Hafner-Burton, Von Stein and Gartzke (2008) and Milner (1998), we opened this paper with what we think to be a fairly simple proposition: scholars can better understand outcomes in IOs by developing theories that explicitly make assumptions about legislative process. Outcomes in IOs are the end of a bargaining process in an institutional setting defined by rules and process—a bargaining process that we need to better understand lest it be corrupted by powerful actors (Vreeland 2018). By way of illustrating the contrast between the study of outcomes and process in international assemblies, consider the following anecdote from the UNGA.

During the General Assembly’s 69th session in 2014, the Social, Humanitarian, and Cultural Committee—otherwise known as the Third Committee—considered a draft resolution addressing the “Situation of Human Rights in the Democratic People’s Republic of Korea (DPRK).” The resolution acknowledged the findings of a Human Rights Council commission of inquiry report that accused the DPRK of gross violations of human rights, and it called for a cessation of the government violence. During committee consideration of the resolution, Cuba proposed an amendment to replace operative paragraphs 7 and 8—which indicated reasonable evidence of DPRK “crimes against humanity” and stated the UNGA’s intention to submit the report to the UNSC and encourage the referral of responsible parties to the International Criminal Court—with much friendlier language. The committee vote on Cuba’s amendment failed 77 to 40; 50 countries abstained. After the failure of the amendment, the Third Committee voted to adopt the original resolution 111 to 19 with 55 abstentions. A month later, the resolution was adopted on the plenary floor by a vote of 116 to 20 with 53 abstentions; there was never a vote on Cuba’s amendment in the plenary.

Absent a consideration of process, IO scholars might view this as one that easily passed on the UNGA plenary floor. None of the disagreement nor procedural jockeying that occurred in the committee process would be reflected in that data point. Even more importantly, none of the disagreement found in the failing amendment vote tally would be reflected in the lopsided vote tally in support of the resolution on the plenary floor. To be clear, we are not calling for a detailed account of every policy fight that occurs in every international assembly. Rather, an investigation grounded in a legislative process-based theory—focusing on committees, voting rules, amendments, and scheduling power—would not only help us to explain this interaction, but would also potentially generate additional hypotheses about power and influence to help us better understand power in the international system.

Footnotes

Notes

Funding for this project was provided in part by the National Science Foundation Political Science Grant (SES #1625064), the Institute for Global Conflict and Cooperation (IGCC), and the Tony Coelho Endowed Chair of Public Policy and the Committee on Research (COR) at the University of California, Merced. Thanks to the Editors at ISR, two anonymous referees, Chuck Finochiarro, and Amy Yuen for thoughtful comments on previous drafts and to Mike Colaresi, Kennan Conner, Christian Fong, Anupam Dhillon, Stephanie Nail, Tessa Provins, Lila Suboh, Erik Voeten, Nick Weller, and Ae sil Woo for valuable discussions on this project. We are grateful to Kennan Conner, Josh Franco, Alison Luna, Stephanie Nail, Tessa Provins, and Ae sil Woo for valuable research assistance.

Associate Professor of Political Science. Email: [email protected]

Professor of Political Science and Tony Coelho Endowed Chair of Public Policy. Email: [email protected]

REFERENCES