Constitution-Building Bodies and the Sequencing of Public Participation A Comparison of Seven Empirical Cases

Constitution-building is one of the most salient aspects of transitional processes, from war to peace or from authoritarian rule, in terms of establishing and strengthening democracy. This paper is part of a research project that aims to identify the circumstances under which constitution-building can strengthen democracy after violent conflict and during transitions from authoritarian rule. Previous research has indicated that the actions and relations of political elites from opposing political parties when making the constitution has bearing on the state of democracy post promulgation, but that the careful sequencing of public participation in the process can be of relevance as well. This paper conducts a systematic analysis of seven empirical cases and focuses the investigation to the type of constitution-building body that has been employed and to during what stage of the process the general public have been invited to participate. It concludes that popularly elected constitution-building bodies tend to include a broad range of political parties and that they, additionally, tend to have rules of procedure that encourage compromise and negotiation, whereas appointed bodies are dominated by one single party or one single person and do not have rules of procedure that necessitate compromise. The paper also discusses the potential need for political elites to have negotiated a number of baseline constitutional principles prior to inviting the general public to get involved in the constitution-building process, and concludes that this is an area of research in need of further in-depth empirical case-studies.


Introduction
Constitution-building is indisputably one of the most salient aspects of transitional processes, from war to peace or from authoritarian rule, in terms of establishing and strengthening democracy (Diamond et al., 2014;Bell, 2016).When former adversaries of violent conflict agree to lay down their arms and engage in peace talks, and when authoritarian rulers reach agreement with the political opposition to steer the country on a democratic path, constitution-building is almost always part of the settlement (Bell & Zulueta-Fülscher, 2016).Hence, constitution-building appears to be an inherent part of transitions; it marks a break with the past and the turning of a new leaf.But even though the importance of constitution-building in post-conflict states and in states transitioning from authoritarian rule is by now widely acknowledged (see e.g.Hart, 2010;Banks, 2008;Ginsburg et al., 2009;Bell & Zulueta-Fülscher, 2016;Miller, 2010;Brandt et al., 2011), much less is known about how constitution-building processes can contribute to strengthening democracy in the aftermath of violent conflict and during transitions from authoritarian rule.
During the past twenty-five years, the conventional view in scholarly circles and in practical field work, has almost unequivocally pointed to broad based public participation in constitution-building processes as the most important and decisive factor in increasing levels of democracy post-promulgation (see e.g.Ghai & Galli, 2006;Samuels, 2009;Ponzio, 2011;IDEA, 2011).In essence; the people are to be heard, and their opinions taken into consideration, when it comes to establishing the principles that will provide the cornerstones of the political rules of the game.
To be sure, public participation in constitution-building holds the potential to give rise to a number of positive effects, not least at an induvial level of analysis (see e.g.Moehler, 2008).Recent work has, however, empirically substantiated that there is no explicit relationship between participatory constitution-building and higher levels of democracy after the finalization of the process and the enactment of the new constitution (Saati, 2015).Rather, the actions of political elites in constitution-building processes and the nature of their relationships appear to matter more for democratization than what has hitherto been acknowledged in this specific strand of research.At the same time, there is reason to further empirically probe if public participation in constitution-building, during certain junctures of a constitution-building process, can be more conducive for strengthening democracy (Saati, 2015).In this paper, I begin to explore how these two, seeming opposites -political elite negotiations and public participation -may be sequenced in the context of constitution-building.The paper pursues this exploration by the comparison of seven empirical cases: Afghanistan, East Timor, Ethiopia, Iraq, Rwanda, Tunisia and South Africa.
In order to develop our understanding of the actions of political elites during the process of constitution-building; how they reach settlements, and how they engage in political negotiations, the institutional structure of constitution-building bodies are crucial to consider (Ginsburg et al., 2009).This since it is in these formal bodies that much of the negotiations take place.Therefore, this paper sets out to systematically investigate the setup of the formal constitution building-bodies that were used in the cases under investigation, and asks: was the constitution-building body broadly inclusive or dominated by one political party?By which rules were decisions made in these formal bodies?Type of constitution-building body has been theorized to matter, but systematic analysis on this topic has up until now remained absent (Ginsburg et al., 2009).
When it comes to public participation in constitution-building processes, in turn, other aspects are relevant to consider.Although participatory constitution-building in post-conflict settings and in states transitioning from authoritarian rule has developed into modus operandi, to the extent that it has become a new norm (Hart, 2010), the practice of it may sometimes be problematic.Additionally, participatory constitution-building also raises a number of questions that the research community still need to systematically investigate and provide answers to (Brandt et al., 2011;Moehler & Marchant, n.d.,;Kant & Rakuita, 2014).Furthermore, broad based public participation in constitution-building is, for a number of practical reasons, not always an actual possibility for many post-conflict states and states in transition from authoritarian rule (Saati, 2015).Therefore, it is relevant to explore if the participation of the public is more conducive during specific stages of the constitution-building process.In the process of systematically analyzing the types of formal constitution-building bodies that have been used in the cases under investigation, this article will therefore also devote attention to exploring during what stage of the process the public have been invited to participate in order to understand if a sequencing pattern can be detected in terms of when it might be more fruitful to involve the public.This paper is organized as such: In the next section I briefly discuss key concepts and definitions.This is followed by a theoretical discussion concerning political elite settlements and negotiations and the link to democratization, as well as public participation in constitution-building and the link to democratization.Subsequent to this I explain the rationale that has guided the case selection and also present the data which has been utilized to bring clarity to the central questions of this study.Thereafter empirical results are presented and discussed.In the final section of the paper I discuss some conclusions and suggestions for future research.

Defining Concepts: Constitution-Building, Democratization and Public Participation
In accordance with previous scholarly work within this strand of research, this study understands the term constitution-building as encompassing all phases involved in the process that results in having a written constitution, whereas constitution-making refers to the actual writing of the text (Ghai & Galli, 2006: 9).This distinction ties into the idea of constitution-building as a process.Hence, the use of the term constitution-building processes signifies that I am concerned with such matters as the procedures through which the constitution is negotiated, the formal bodies used to arrive at a final document, the ways in which public views are allowed to be expressed, how decisions about the content of the final document are made, etc.
For the purposes of this study, democracy is understood in its liberal form, meaning that I am concerned with the conventional aspects of democracy; the holding of free and fair elections, universal suffrage and in addition to these procedural aspects, a number of political and civil rights (Teorell, 2010: 30-31).Democratization, in turn, can either be understood as the replacement of an undemocratic system with a democratic system or as the process during which a state becomes more democratic -however small that change may be.These two understandings are not the same thing expressed in two different ways.The first interpretation of democratization leans on a dichotomous conceptualization of democracy (democratic/non-democratic), while the second one rests on a graded approach (more democratic/less democratic) (see e.g.Sartori, 1987;Dahl, 1989;Collier & Adcock, 1999;Alvarez et al., 1996;Bollen & Jackman, 1989;Teorell, 2010).In this study, I lean on the later conceptualization.Thus, when referring to democracy enhancement as well as discussing democratization in a country, the issue under consideration is whether a country has become more democratic when comparing its status between two time intervals.
When referring to public participation in constitution-building, it is relevant to understand that, with a historic perspective, this is a new occurrence.Traditionally, constitution-making has been a reserved area for political elites and lawyers (Hart, 2003).In so far as people have been invited to participate, their participation has most often been limited to voting for members of a constitutional assembly; a formal body that is responsible for drafting the document.However, as of the mid 1980s and early 1990s, constitution-building has taken a different turn, especially in post-conflict states, placing ordinary citizens at the front and center of attention (Khanal, 2014: 4-6).Today, public participation in constitution-building involves more direct and active forms of participation.Many times the public are, for example, allowed and encouraged to formulate constitutional submissions for consideration by formal constitution drafting bodies.Public participation also entails the participation of everyone who is actually interested in participating, and previously disenfranchised groups are particularly encouraged to voice their opinions about what they believe the constitution should include (Khanal, 2014: 4-6).

Political Elites, Public Participation and Democratization
The field of research into which this study is situated is interesting for many reasons, one of which is because it engages, and builds its hypotheses, from assumptions derived from two different literatures: writings concerned with democratization theory on the one hand, and the peacebuilding literature on the other hand.This holds the promise of bringing strength and nuance, but it also poses certain challenges -mainly in the sense that there are some incoherence's between these two strands when it comes to how democratization is perceived to be brought about and strengthened.
Within the broader field of democratization theory, there is a specific scholarly strand that places human agency at the center of consideration.These analysists consistently stress the freedom individuals have to make choices, and accordingly they direct attention to the actions that individuals take and the consequences that these actions have for making democratization possible (Rustow, 1970;Bratton & van de Walle, 1997).Events that lead to democratization unfold in ways that are utterly connected to human activity, with the choice of one individual affecting the choices of another individual and so forth.Whether individuals choose to side with the incumbent regime or the opposition depends on subjective perceptions and calculations as to which alternative better suits their own interest.Hence, in order to better understand variations in political outcomes, these scholars suggest that we are well advised to analyze the motivations, preferences and calculations of self-interested actors (Rustow, 1970;O'Donnell and Schmitter, 1986).Larry Diamond, who is a prominent democratization scholar, is so sure about the role of actors for democratization that he has stated: "In my view, the only absolute precondition for achieving a democratic transition, aside from Rustow's background condition of a reasonably coherent state, is a set of elites who decide for whatever reason that democracy is in their interest".(Diamond et al., 2014: 91) Without a doubt, a rational choice logic constitutes the backbone of this actor-oriented analysis as to the causes of democratization (Bratton & van de Walle, 1997: 24-25).As theorized by Dankwart Rustow (1970), the process of democratization springs from interaction and bargaining, particularly among political elites.The surest way towards democracy in this game of bargaining is to make certain that all stakeholders get at least some of their demands met, so that no one is left empty handed (Rustow, 1970;Bratton & van de Walle, 1997;McFaul, 2002).Taking this line of reasoning one step further, O' Donnell and Schmitter (1986) have argued that democratization is determined by interactions and bargaining, between, on the one hand, hardliners and soft-liners who support the incumbent regime, and, on the other, between the radical and moderate segments of the opposition who do not support it.
The emphasis on political elites, their ability to negotiate, bargain and reach settlements, has also been brought to attention in research on constitution-building in post-conflict and transitioning contexts (Bell & Zulueta-Fülscher, 2016;Bell, 2016).In parallel to this, a burgeoning academic interest for the actions of political elites suggests that consensus politics and bargaining between politicians from opposition factions in the context of constitutional negotiations are key in explaining positive developments in a number of empirical cases as far as democratization post-promulgation of the new constitution is concerned (see e.g.Cheeseman & Tendi, 2010;Filali-Ansary, 2015).The challenge of properly balancing trade-offs between short term requirements, in terms of getting all relevant stakeholders to agree on a political settlement, and long term requirements of securing long lasting peace, stability and democratization is being increasingly problematized in this vein of research (see e.g.Brandt et al., 2011;Bell & Zulueta-Fülscher, 2016) which has, at times, been accused of providing one dimensional solutions that do not resonate with practical realities in many post-conflict contexts (cf.Richmond & Mac Ginty, 2015) -participatory constitution-building being one of these suggested solutions.It is, however, increasingly being acknowledged that public participation in constitution-building is not a cure-all solution for many of the challenges faced by post-conflict states and states transitioning from authoritarian rule (Saati, 2015;Bell & Zulueta-Fülscher, 2016).As amply put by Bell: "…those who seek to rely on constitutions for conflict resolution, need to understand this enterprise as just a political and fraught as all other institution-building efforts".(Bell, 2016: 4) This could be interpreted as the need for political groundwork in terms of some form of political settlement to be established as a prior condition before, basically, anything else.In other words; if there is no political elite buy-in to begin with, if political elites cannot consensually agree on the political rules of the game, there is very little public participation in the constitution-building process can achieve in terms of democracy enhancement at the macro level of analysis (Bell & Zulueta-Fülscher, 2016;Bell, 2016).It is against this backdrop that the composition of the formal constitution-building bodies that have been employed in post-conflict states and in states transitioning from authoritarianism becomes a pertinent area of inquiry.However, for the purposes of this paper it is also relevant to understand how advocates of participatory constitution-building theorize that participation leads to democracy enhancement at the macro level of analysis.This is the focus of attention in the following section.
As mentioned earlier, it is not only democratization scholars who are interested in understanding how constitution-building in post-conflict states and in states transitioning from authoritarian rule can contribute to enhancing democracy.This research interest is also shared with peacebuilding scholars, as well as peacebuilding practitioners (see e.g.Samuels, 2009;Ghai & Galli, 2006;IDEA, 2011;IDEA, 2012).When reviewing the arguments in favor of public participation in constitution-building presented by peacebuilding scholars, it becomes quite clear that the post-conflict state of being is, in essence, the core upon which their theoretical claims are being built.Indeed, as post-conflict states, and states in transition from authoritarianism, are in short supply of social capital as well as democratically functioning institutions, measures that establish both of these must be taken, if the state of democracy is to have a viable chance of being strengthened (Moehler, 2008).Towards this end, the active participation of the general public is viewed as absolutely indispensable.In order to build mutual trust between citizens, and between citizens and the government, so that social capital is reinforced, the rules upon which the political system are built must be perceived as legitimate (Moehler, 2008;Widner, 2005;Widner, 2008;Hart, 2003).For these rules to be recognized as legitimate, in turn, people must be allowed to be part of formulating those (Hart, 2003).In essence, it could be said that supporters of participatory constitution-building, and of direct citizen engagement in issues of public relevance in general (see e.g.Pateman, 1970), envision a self-enforcing loop in which participation in constitution-building is held to bolster democratic attitudes, in turn making people more open to listening and respecting the views of others, in turn leading to the development of a democratic political culture, leading to state institutions being strengthened (Moehler, 2008: 33).
To summarize this part of the paper; what we have are two strands of literature, both of which are interested in understanding how constitution-building in post-conflict states and in states transitioning from authoritarianism can strengthen democracy, but they approach the issue from almost diametrically opposed theoretical perspectives.Whereas many democratization scholars focus on contingent factors and place emphasis on human agency, and particularly the actions of political elites for making (or breaking) democracy, peacebuilding scholars and practitioners stress the role of the general public in the constitution-building process as vital for long term democratization.My argument here is that these two need not negate each other, but that they could rather be combined, and perhaps sequenced.

Methodology
Before continuing to the empirical section of this paper, some notes on methodology are warranted.The empirical work presented here is a qualitative comparative analysis in which seven cases are systematically compared and analyzed, specifically in regards to the formal constitution-building bodies that were used in each of the cases as well as in regards to during what stage of the process the general public were invited to participate.As the number of cases extends beyond two or three, the analysis sets out to explore if any patterns can be detected as to whether or not the set-up of these constitution-building bodies matter when it comes to facilitating political elite cooperation.Increasing the number of cases beyond very few will hopefully also shed some light on whether there are certain stages of the constitution-building process that are more conducive for involving the general public.
When it comes to material, the main source is written documentation.Previous case study research on constitution-building processes in different locations are helpful as they explain the context in which the process was instigated, the type of formal bodies that were used, the extent and type of constitutional education programs that were made available for the public, etc.Just to name a few of these previous studies which have been useful for the work presented in this paper; Brandt's (2005) contribution on the East Timorese process and the process in Afghanistan, Morrow's (2010) contribution on the Iraqi case, Ebrahim and Miller's (2010) insights as regards the South African process, and Wodajo's (2001) expertise on the Ethiopian process are all valuable contributions that this study has found most useful.In some instances, when constitution-building has taken place in the aftermath of war, guidelines concerning types of constitution-building bodies have been formalized in official peace agreements -in such cases, these peace agreements are also part of the written documentation that has been analyzed (for example, the Arusha Accords that were signed by the warring parties in Rwanda to end the civil war there in 1993).
As regards the selection of cases, this is based on certain pre-determined criteria.Attention is devoted to 1) constitution-building processes after violent conflict and during transitions (Note 1) from authoritarian rule in which 2) the public have been invited to participate, and where 3) democracy levels have increased after the enactment of the new constitution.The temporal scope of the investigation is limited to constitution-building processes that have taken place after the end of the Cold War (i.e. in the period 1989-2016) as it is from the onset of the 1990s and the ushering in of the peacebuilding era that participatory constitution-building has become more common.Seven cases meet these criteria and are included in the study.These cases are: Afghanistan, East Timor, Ethiopia, Iraq, Rwanda, Tunisia and South Africa.The criteria of case selection also makes it clear to the reader that cases are selected on the dependent variable (only cases in which democracy levels have increased are included in this study).The rationale for this methodological choice ties into the overarching purpose of this investigation: to identify circumstances and conditions under which constitution-building processes lead to higher levels of democracy.Therefore, cases with a negative outcome as far as democracy levels after the enactment of the constitution is concerned are not relevant to include in order to fulfill this purpose (George & Bennet, 2005: 23).Indeed, had the purpose been different, cases in which democracy levels have declined following the enactment of the constitution would have been included as well, noticeably increasing the number of cases.
In order to determine whether democracy levels indeed have increased following the enactment of the new constitution, empirical indices from Freedom House and the Polity IV project have been probed (Freedom House, Freedom in the World Ratings 1972-2016;Polity IV Project, Political Regime Characteristics andTransitions, 1800-2015).Democratic status has been compared, for each of the seven cases, five years prior to the commencement of the constitution-building process with the democratic status after the first elections following the enactment of the new constitution.As the Polity IV project suffers from missing data for some of the cases, the main source of data has therefore been retrieved from Freedom House, but in instances where data from Polity IV has been available, this data has been crosschecked with the Freedom Houses data.In Table 1, Freedom House's data for the seven cases are displayed, five years prior to the commencement of the constitution-building process compared to the score for each of the cases following the first elections after the finalization of the process and the enactment of the new constitution (Freedom House, Freedom in the World Ratings 1972-2016).Please note that Freedom House's scale, ranging from 1-7, is reversed in the sense that a lower score implies greater political rights and civil liberties than a higher score, which indicates an unfavorable situation as regards political rights and civil liberties.The table shows that in these seven cases, democracy scores have indeed increased following the enactment of the new constitution -in some more so than in others, with the South African and Tunisian cases showing the most positive notable change between the two intervals.A few words about empirical indices for measuring democracy are also warranted.To be sure, there is an entire scholarly field dedicated to analyzing and evaluating such indices (see e.g.Norris, 2008;Hadenius & Teorell, 2005;Munck & Verkuilen, 2002) and this is not the place to give an account of this literature.Two points are however relevant to raise.The first point is that in the scholarly field of comparative democratization, using Freedom House's indices and the Polity IV data is standard procedure (Norris, 2012: 14;Bogaards, 2012: 691;Hadenius & Teorell, 2005: 26), which motivates using these indices in this study as well.The second point is that different empirical indices of democracy, even those that adopt a minimalist definition of democracy and conceptualize democracy in a dichotomous way (democracy/dictatorship), correlate strongly and significantly with both Freedom House's index of democracy and the Polity IV indices, both of which include a maximalist definition of democracy in the sense that civil rights are also included in the measure (Norris, 2008: 71).Even more relevant, for the purposes of this study, is that these two empirical indices of democracy correlate more than any other of the empirical indices available; with an N of 4382 cases, the correlation between Freedom House's measures and the Polity IV project is .904(Norris, 2008: 71).

What Type Of Constitution-Building Bodies Were Employed?
In this part of the paper, I begin with a systematic empirical investigation concerning the type of constitution-building body that was used in the seven cases in order to understand if the body in question facilitated political compromise between different factions.In the section that ensues (5.2), the issue of sequencing public participation into the process is the focus of the empirical investigation.

Type of Constitution-Building Bodies in Seven Cases
Though type of constitution-building body has been theorized to matter (Ginsburg et al., 2009), a review of the empirical material makes it relevant to ponder if it is perhaps not as much the type per se that matters, but rather the issue of whether or not the body in question has been an inclusive or an exclusive one.The empirical material displays a broad variety of types of constitution-building bodies that have been used in the different cases, and also shows that most cases have employed a number of different constitution-building bodies during the course of the process and that these, in turn, have had different mandates.Let us turn to the cases under investigation.
In Afghanistan a constitutional commission was used (Brandt 2005), whereas in South Africa, a constitutional assembly was employed (Ebrahim & Miller, 2010).Constituent assemblies were the constitution-building bodies of choice in East Timor and Ethiopia (Brandt, 2005;Wodajo, 2001), and constitutional drafting committees in Afghanistan (in addition to the constitutional commission used there) and Iraq (Brandt, 2005;Morrow, 2010).In Ethiopia a constitutional drafting commission was employed in addition to the constituent assembly (Wodajo, 2001).In Rwanda, a legal and constitutional commission was used as was a transitional national assembly (Banks, 2008; The Arusha Accords), whereas Tunisia used a national constituent assembly and a national dialogue (The Carter Center, 2014).Finally, traditional constitution-building bodies, such as the constitutional Loya Jirga have also been employed; in Afghanistan (Brandt, 2005).
When studying the constitution-building bodies that were used in each of these cases, a distinction can indeed be made between cases that have used inclusive constitution-building bodies and those that have used exclusive constitution-building bodies.Inclusive in this sense signifies constitution-building bodies that have included a broad spectra of different political parties rather than having been dominated by one single party.Exclusive, on the other hand, signifies the reversed; i.e. that the constitution-building body in charge of drafting and/or approving the constitution is dominated by one single political party, or even by one single person (usually the executive), or that certain political parties have been altogether disqualified from engaging in the process.In Table 2, the seven cases are organized according to whether inclusive or exclusive constitution-building bodies were employed, as well as whether these were popularly elected or not.Turning to the left hand column we can see that East Timor, Ethiopia, Tunisia and South Africa are cases that have used inclusive constitution-building bodies.It should nevertheless be noted that although the constitution-building body in charge of drafting and/or adopting the constitution was inclusive, in three of these cases -South Africa, Tunisia, and East Timor -one political party managed to acquire a majority of the seats in the body (Ebrahim & Miller, 2010;The Carter Center, 2014;Brandt, 2005).This could, in turn, indicate that the constitution-building process was not as inclusive as one might believe.Interestingly, however, in these cases, the rules of procedure for reaching decisions about the content of the constitution, or for approving the constitution in its entirety, was either by two-thirds majority or by consensus.
Such rules of procedure implied that in Tunisia, where the moderate Islamic political party Ennahda secured the majority of seats in the National Constitutional Assembly (NCA) -the body in charge of drafting the constitution -members of this party still had to find common ground and compromise with members from secular political parties in order to acquire the needed support for their constitutional suggestions (The Carter Center, 2014).This need for compromise was, furthermore, absolutely necessary for the drafting exercise to be able to proceed.The Carter Center's analysis of the Tunisian process captures this mechanism for reaching compromised agreements quite well.They state that: "the strength of the Tunisian model lies in the tireless work of the NCA and the political parties to overcome divisions and build consensus, resulting in the adoption of a constitution backed by the vast majority of deputies in an otherwise highly divided assembly".(The Carter Center, 2014: 7).
In South Africa, as well, the African National Congress (ANC) held a majority of the seats in the Constitutional Assembly but due to the two-thirds majority requirement for adopting the constitution, members of this party also had to compromise with colleagues from other political parties, most notably with the National Party (NP) (Ebrahim & Miller, 2010).And lastly, Fretilin in East Timor; Fretilin held the majority of seats in the Constituent Assembly but also had to find support for its constitutional suggestions among oppositional parties due to the rules of procedure for adopting the constitution that stated that at least 60 representatives of the 88 had to vote in favor of it (Brandt, 2005).
It is worthwhile to note that the formal constitution-building bodies in charge of drafting and/or adopting the constitution in the four cases accounted for above were popularly elected.This is interesting because it stands in direct contrast to the constitution-building bodies that were employed in the three remaining cases (see the middle and right hand columns in Table 2).In Afghanistan, Rwanda and Iraq, constitution-building bodies were exclusive in either one of the following ways; one single political party dominated the formal body; the executive dominated the body/bodies; or certain political parties/groups were disqualified from engaging in the proceedings of these bodies.
In Afghanistan, it was the president who appointed the nine members of the Constitutional Drafting Committee who were, as the name suggests, in charge of the actual drafting of the constitution.The president also appointed the 35 members of the Constitutional Commission, as well as 50 out of the 500 members of the Constitutional Loya Jirga.When it comes to authority over the content of the final draft, this too was vested in the hands of the president and his cabinet (Brandt, 2005).Adding to what appears to be a rather authoritarian constitution-building process dressed up in a more inclusive manner than what was the actual case, representatives from the former Taliban regime were not allowed to take part in any of the procedures in either one of the three different constitution-building bodies used in Afghanistan (Brandt, 2005).The Rwandese constitution-building body in charge of drafting was likewise an exclusive body in regards to two different aspects.To begin with, the twelve members of the Legal and Constitutional Commission were appointed by the members of the Transitional National Assembly -hence not popularly elected.Secondly, none of the political forces that had taken part in the genocide (mainly Hutu political parties) were invited to participate in the work of the commission (Ankut, 2005: 15-16).In Iraq, at least the members of the Constitutional Drafting Committee were drawn from the popularly elected National Assembly which indicates that there was a window of opportunity for an inclusive constitution-building body.This was however never realized since the Sunni Arab community, the majority population of the country, boycotted the 2005 elections for the National Assembly; hence there were no Sunni Arab members for the Constitutional Drafting Committee to draw from (Morrow, 2010).
The Ethiopian case, lastly, deserves some mentioning as well.In a sense, this case employed a mixed approach of inclusive and exclusive constitution-building bodies.Whereas the Constitutional Drafting Commission was appointed by a Council of Representatives, the Constituent Assembly which had final authority as regards passing/rejecting the constitution, was popularly elected.And even though the Constitutional Drafting Commission was appointed, it was still inclusive in the sense that it included a broad spectra of political and ethnic groups (Wodajo, 2001).Furthermore, decisions in the drafting commission had to be made in consensus (Wodajo, 2001).On the other hand, one of the most important actors in the Ethiopian process, the Oromo Liberation Front (OLF), withdrew its participation from the process during the early stages of its commencement, making the constitution-building process not entirely inclusive after all.
Hence, when one takes all seven cases into consideration, something resembling a pattern is detected; inclusive constitution-building bodies tend to be popularly elected and have rules of procedure for reaching decision about the content of the constitution, and for the adoption of the document in its entirety, that encourage compromise between politicians from different political affiliations.Exclusive constitution-building bodies, on the other hand, tend not to be popularly elected but rather appointed and do not have rules of procedure that require, nor encourage, the reaching of compromise between political opponents.

Involving the Public in the Process
Just as the empirical cases included in the study display a variation as to the type/types of constitution-building bodies that have been used, the cases also differ when it comes to during what stage of the process the public have been invited to participate.In East Timor and Rwanda, the viewpoints of the general public were probed prior to, as well as after the drafting of the constitution (Brandt, 2005;Banks, 2008).In Ethiopia, the public were likewise invited to make constitutional submission prior to drafting began and continuously during the process of drafting but not to provide further input on the draft once it was finalized (Wodajo, 2001).In Iraq, public participation was limited to the drafting phase of the process -hence, public input did not provide the platform upon which the constitutional drafting committee constructed the constitution (Morrow, 2010), whereas in Afghanistan, Tunisia and South Africa, the public were invited to participate after the formulation of the draft (Brandt, 2005;The Carter Center, 2014;Ebrahim & Miller, 2010).There are however some important differences between these cases.In particular, the South African case stands out from the rest.
In light of the focus of this paper, the South African constitution-building process offers some interesting insights that speak to the issue of sequencing public participation into the process.That this particular case of constitution-building is frequently brought to attention for its highly participatory features is quite known by scholars as well as constitution-building practitioners (see e.g.Democracy Reporting International, 2011; Ebrahim & Miller, 2010;Haysom, 2001).What is not analyzed quite as often, however, is how the participatory phase fitted into the rest of the process.In fact, when one analyzes the South African constitution-building process, it appears that this process was an exercise in carefully sequencing events, in which the first stage of the process was completely limited to political elite negotiations behind closed doors, far from the public eye (Mbete-Kgositsile, 2001).Once the South African political elites had reached consensus on a number of contentious constitutional issues -34 constitutional principles to be precise -the South African population at large was invited to participate in the constitution-building process (Ebrahim & Miller, 2010).Hence, there was a negotiated platform, agreed upon by political elites from a broad spectra of political parties, which preceded the public participation phase of the constitution-building process.
It is likely that this specific sequencing pattern had two major effects: for one, the agreed upon platform containing the 34 constitutional principles, may have contributed to keeping the input of the general public focused, as the public were aware that constitutional submissions contradicting the already agreed upon principles would not be taken into consideration by the constitutional drafters.Second, by being aware that there were some baseline principles that had already been consensually agreed on by the political elites of the country, the general public's expectations as to having all of their constitutional suggestions reflected in the final document was circumvented, which in turn may have lowered the risk of public frustration and possible violent outbreaks.Another, yet related aspect, that also plays into the issues of sequencing, keeping the discussions focused, and managing people's expectations which is also particular for the South African process, concerns the thoroughness and the extent of constitutional education programs that were made available for the public prior to soliciting their views.These constitutional education programs made it possible for South African's to not only focus their suggestions to areas that were relevant to consider in a constitution, but to also make informed choices based on different alternatives (Ebrahim & Miller, 2010).
The Tunisian case also had a similar set-up as the South African case, i.e. a two-pronged constitution-building process in which the first phase was limited to political elite negotiations between, in particular, members from Ennahda and members of secular political parties.The way this two-pronged process came about, however, was rather due to organizational failure than to a planned sequencing arrangement as the one in South Africa (The Carter Center, 2014: 68).Though there was a determination to include the Tunisian citizenry in the making of the constitution -the rules of procedure for the process even stated that the members of the NCA were to directly reach out to, and engage with, the Tunisian public (NCA, Rule of Procedure, Rule 79, 2011) -logistical, financial and administrative failures got in the way of this actually materializing, and therefore public consultations did not take place until the second draft of the constitution had been finalized (The Carter Center, 2014: 69).Seeing how the Tunisian process included the formulation of no less than four drafts, the Tunisian citizenry were hence involved in the constitution-building process but not until half way through.In turn, though it was not planned for, political elites were able to negotiate and compromise on the sensitive issue of how to reconcile the issue of giving the country a specific religious identity whilst also implementing democratic rules (Filali-Ansary, 2014).
Though we can of course never know for sure how events would have developed had the NCA in Tunisia been able to act accordingly to its rules of procedure and include the Tunisian citizenry in the process from the very early stages, in hindsight it could be argued that keeping the first phases of the process limited to political elite negotiations was a successful approach in order to reach compromise and keep discussion from becoming too polarized; something that is actually one of the potential risks of participatory constitution-building (Gluck & Brandt, 2015: 15).Indeed, as suggested by Elster (1995: 388), though public debate and input has its benefits, it can also lead to political elites grandstanding and overbidding each other in order not to be viewed as "sell outs" by their constituents, while private negotiations amongst elites, on the other hand, tend to provide more favorable circumstances for reaching compromise.
As far as Afghanistan is concerned, the authoritarian structure of the constitution-building process had, perhaps unsurprisingly, an effect on the public participation phase as well.Citizens were not invited to participate until the Constitutional Drafting Committee had already produced a draft at which point the Constitutional Commission had a mere month at its disposal to educate the entire population on the content of the draft as well as perform consultations to solicit submissions (Brandt, 2005).Poor infrastructure, a high illiteracy rate, and vast obstacles of reaching women and the disabled made it impossible to manage the task of constitutional education within the time frame, in turn making it difficult for people to be able to genuinely participate (Saati, 2015).On the issue of timeframes, the empirical material also reveals that cases in which international actors have had an especially strong presence -East Timor, Iraq and, again, Afghanistan -have had constitution-building processes that have been characterized by extremely short timeframes (Brandt, 2005;Morrow, 2010).The East Timorese constitution-building process was to be finalized within 90 days, the Iraqi process within 9 months, and constitution-building in Afghanistan within 18 months (Brandt, 2005;Morrow, 2010).Considering the major obstacles for engaging ordinary citizens in the processes of these cases, in terms of insufficient infrastructure,technical as well as geographical -little or no culture of constitutional democracy, low levels of education among the citizenry, etc., it is not surprising that public "participation" was never genuinely achieved (Saati, 2015).Paradoxically, it is often times international actors who push domestic actors to involve their citizens in the making of the constitution (Bell, 2016), while at the same time the same international actors appear to be reluctant to allow the acquired time for genuine participation to be able to be achieved.
As regards the other cases -Ethiopia and Rwanda -both involved their respective populations in a similar manner with the one main distinction being that in Ethiopia public input on the finalized draft was not part of the process whereas it was in Rwanda.What also sets these two cases apart from most from the other cases but brings them closer to the South African experience, was the extent of constitutional education programs offered to the general public during the entire course of the process.These measures, ranging from TV shows, radio programs, puppet shows, essay contests, etc., gave rise to an informed Ethiopian and Rwandese population who were adequately educated on constitutional issues to be able participate in the constitution-building process of their countries in a meaningful way (Wodajo, 2001;Ankut, 2005).
To conclude, when it comes to public participation in constitution-building processes and the issue of sequencing, it is a bit more challenging to draw definite conclusions and ascertain clear patterns in the data that point in one clear direction as to when during the process it might be "best" to involve the public.However, in light of the South African and Tunisian experiences, there is reason to further investigate and discuss the potential benefits of employing a two-pronged set-up in which a broad spectra of political elites negotiate the most sensitive constitutional issues during the first stage of the process after which the general public are invited to provide input during the coming stage/stages of the drafting exercise.

Concluding Remarks
The work presented in this paper is an initial exploration of how two, seeming opposites -political elite negotiations and public participation -can be sequenced in the context of constitution-building based on the premise that both of these aspects can be important for democracy enhancement in post-conflict states and in states transitioning from authoritarian rule.Previous research has established that political elite negotiations and political settlements are more important than what has thus far been acknowledged in this scholarly field.Therefore, it is relevant to analyze how such negotiations can be facilitated, and particularly so in formal constitution-building bodies.In light of how public participation in constitution-building processes, particularly in post-conflict states, has developed into a new norm, it is however equally important to also understand how public involvement can be sequenced into the broader process in a manner that is conducive for long term democratization.
This study has provided a systematic analysis of the types of constitution-building bodies that have been used in seven empirical cases -seven cases in which democracy levels have indeed increased following the enactment of the new constitution.Having done so, it concludes that something resembling a pattern can be detected in terms of popularly elected bodies being more inclusive than appointed bodies and also in terms of having rules of procedure for reaching decisions that encourage compromise, dialogue and negotiations between political opponents.If the current trend of participatory constitution-building in post-conflict states and in states transitioning from authoritarian rule endures, we are bound to witness an increase of such cases in the future.Therefore, forthcoming research should aim to further test and validate this finding by including a larger set of cases in the analysis.At the same time in-depth case-study research on the internal proceedings of formal constitution-building bodies is an area of study that ought to be pursued so that we may learn more about how, and why, political elites compromise and find common ground when it comes to contentious constitutional issues.How do these different stakeholders make sure that no one loses out in this game of bargaining while at the same time a coherent and democratic constitution is produced?This is an area of inquiry in dire need of scholarly attention.
This paper has also mapped out during what stage/stages of the constitution-building process the general public in the seven empirical cases were invited to engage in the making of their constitution.In regards to this issue, it is more difficult to detect any clear pattern or draw definite conclusions as to during what stage of the process it might be "best" to include the public, and a prescriptive blue-print might not be desirable either.Nevertheless, some ideas can be tentatively elaborated.Though it is not possible to make causal inferences it is still interesting to note that the two cases -Tunisia and South Africa -that have improved their democracy scores the most when comparing levels prior to the commencement of the constitution-building process with levels following the enactment of the constitution, are the two cases that employed a two-pronged process in which the first stage was strictly limited to political elite negotiations in formal constitution-building bodies that necessitated compromise, negotiations and bargaining.This finding supports the whole idea of carefully sequencing public participation into the broader process of building the constitution, and it also supports the notion of the importance of political elite cooperation as an all but essential stepping stone for the subsequent engagement of the public at large.I would even argue that the findings of the South African and Tunisian cases are sufficiently strong to motivate further case-study research; case-study research that includes in-depth interviews with South African and Tunisian political elites who were part of negotiating the constitutions in these two countries, as this will enhance our understanding of if, and if so how, some of the main stakeholders viewed the constitution-building process, the ways through which they believed that the citizens of their countries were to get involved and how long term democratization could be achieved.

Table 2 .
Inclusive or exclusive constitution-building bodies