Using Evidence in a Highly Fragmented Legislature: The Case of Colombia’s Health System Reform

This chapter examines how evidence is used in major policy health policy initiatives in a highly contested political context. Through a case study of legislation proposed in the context of Colombia’s ongoing health systems reformed process, it explores how such use is affected by the specific role played by the legislature within a highly fragmented polity. We use an institutionalist framework to identify three concentric layers of fragmentation: at the social, political and administrative levels. The former refers to macro levels social structures and factors shaping Colombian society and politics, including the ongoing armed conflict as associated social cleavages which have loomed over Colombian society for decades. At the second level, Colombian politics is characterised by deep divisions and political cleavages along party lines, coupled with weak party structures. This results in a highly fluid political terrain in which new parties may

quickly emerge and disintegrate, and are often held together by 'political strongmen' around which actors coalesce. At the administrative level, the weakness of the legislative and executive branches and corruption endemic in Colombian politics lead to further fragmentation and inefficiency in decision making with the judiciary stepping into the power vacuum to address the most pressing health systems issues (Hawkins and Alvarez Rosete 2017).
The role of the Colombian legislature in the health policy process has to be understood against this challenging backdrop of its recent political and societal history. In the last decade, Colombia's democracy began the long and difficult process of addressing and moving beyond deeply embedded political and societal conflicts in the form of terrorism, internal armed conflict, the illicit drug trade, clientelism and political corruption, which collectively led to the "partial collapse" of the state in the late 1980s (Bejarano and Pizarro 2002). This was reflected in Colombia's low ranking in the World Bank's Worldwide Governance Indicators (WGIs) project on "Political Stability and Absence of Violence/Terrorism", "Rule of Law" and "Voice and Accountability". 1 However, in September 2016, the World Bank president Jim Yong Kim welcomed advances in the peace process aimed at ending the internal armed conflicts, stating that "the country is closer than ever to putting an end to this vicious cycle, and to starting the long and challenging process of transformation and territorial development" (Kim 2016). 1 The Worldwide Governance Indicators project reports aggregate and individual governance indicators for over 200 countries over the period 1996-2015, for six dimensions of governance: Voice and Accountability; Political Stability and Absence of Violence/Terrorism; Government Effectiveness; Regulatory Quality; Rule of Law; and Control of Corruption.
"Political Stability and Absence of Violence/Terrorism" captures perceptions of the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including politically-motivated violence and terrorism. In this dimension, Colombia scored a percentile rank of: 8.2 (1996), 0.97 (2003), 12.3 (2011) and 12.38 (2015).
"Rule of Law" captures perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. In this dimension, Colombia scored a percentile rank of: 22.01 (1996), 26.32 (2003), 47.42 (2011) and 44.71 (2015).
"Voice and Accountability" captures perceptions of the extent to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media. Percentile ranks indicate the country's rank among all countries covered by the aggregate indicator, with 0 corresponding to lowest rank, and 100 to highest rank. In "Voice and Accountability", Colombia scored a percentile rank of: 29.33 (1996), 34.13 (2003), 46.48 (2011) and 45.81 (2015).
See http://info.worldbank.org/governance/wgi/#reports Despite these improvements in recent years, analysts have argued that the various threats faced by Colombia have undermined the legitimacy of state institutions. This in turn can explain much of the current configuration of Colombian politics and the poor institutional performance of key bodies, including the legislature, which has been characterised as highly fragmented and decentralized (Pachón and Johnson 2016;Botero et al. 2010;Bejarano and Pizarro 2002).
The chapter examines evidence use in a fragmented and conflict-riven political environment such as Colombia, focussing on the long-standing and highly politicised attempts to reform the Colombian health system. It shows that policy relevant evidence has consistently been used to inform and provide the rationale for draft laws submitted to Congress over the period of the reforms and appropriate and robust research had indeed entered onto policy making agenda and was cited in the legislation examined. However, reflecting the role of the legislature in the highly contested political system and the health policy subsystem, evidence was not able to change actors´ initial positions and opinions and thus, political consensus on the direction of reforms could not be forged on the basis of the evidence cited. Thus while the importance of evidence to both the substance and politics of the proposed reforms was acknowledged, the inability of policy relevant evidence to overcome political divisions and form the basis of political compromise and consensus is also clear. The analysis presented here supports the conclusion that, in highly contested and fragmented political environments, evidence tends to be secondary to other political and ideological factors influencing policy change. The chapter focuses on the reform process of the Colombian health system and the uptake of research within draft laws submitted to the Colombian Parliament between 1993 and 2016.

Methodology
This chapter draws on data generated from an analysis of draft laws which are available in the public domain as well as from data gathered from semistructured interviews with policy actors in Colombia. As in many countries, Colombia has a hierarchical legal structure with the constitution on top, followed by laws produced by the Congress, which can be Statutory laws (leyes estatutarias), Organic laws (leyes orgánicas) and Ordinary laws (leyes ordinarias) (Vanegas 2012). Any draft law in Colombia has to have an introductory preamble (exposición de motivos) which explains the nature and scope of the problem that it aims to tackle, reviews the regulatory gap that it aims to fill and presents the broad content of the law. It is in the introductory preamble of draft laws where research, if it has been reviewed to formulate the policy proposal, will be referred to. Thus, an analysis of the evidence cited in the introductory preambles allows us to examine whether research evidence has been at all up taken during the policy formulation phase and the body of research which had informed the wider policy debates.
As detailed in Annex 1, 62 draft laws were identified and selected which were then reviewed independently by the present authors of this chapter to assess whether they cited evidence to sustain their policy proposals and, if so, which type of evidence was considered. The criteria used for this assessment was whether the draft law included: (i) data produced by government and its agencies and/or academic studies to consistently define the nature and scope of the problem; and/or (ii) existing national and international research (i.e. international organizations reports, research published in peer-review journals, etc.) to sustain policy proposals. This research, however, did not seek to evaluate the "quality" of the evidence considered and eventually taken up in the draft laws.
We conducted a total of 26 interviews in Colombia in February 2014. Respondents included policy advisers and civil servants at the national level, interest groups representatives, academics, health policy experts and commentators. Through the interviews, we sought to understand the structure and of recent health policy debates in the country and the type of evidence discussed within the policy making processes. In particular, we sought to identify the factors and conditions which helped or hindered the use of evidence to inform those decisions. Interview responses were triangulated with the analysis of draft laws described above and a wider review of relevant policy reports and government documents. To ensure anonymity of respondents we refer to interviews by number. Where it is essential to the understanding or evidentiary weight of quotations, the sector from which respondents came will be detailed. Quotes in Spanish from the interviews and from other bibliographical sources were translated into English by the authors.
the role of the legIslature In coloMbIa The current institutional configuration of the Colombian state is defined by the Constitution of 1991. Colombia is a presidential system in which the President of the Republic is elected directly by the citizens for a set period of four years. The Colombian Parliament is formed of a bicameral Congress with a Senate (Senado), and a Chamber of Representatives (Cámara de Representantes), elected also for a four year period via a proportional representation system. The political party system is weak: parties do not have strong bureaucracies and structures, and are dependent instead upon individual leaders who act as figureheads, tying together otherwise loosely connected political allegiances. Under the current Constitution, the two-party system that had dominated politics prior to the 1990s was replaced by a highly fragmented, multi-party system, with later reforms, in 2003, aimed at reducing this fragmentation. The present system is characterised not only by weak party structures (with parties serving as electoral vehicles for candidates to promote their own personal agendas), but by high electoral volatility, whereby new parties frequently emerge but often cease to exist in one or two elections cycles (Botero et al. 2011;Milanese 2011;Pachón and Johnson 2016).
Within this institutional architecture, what is the role of the legislature in the policy process in Colombia? Saiegh (2010) has suggested a number of factors that can drive a legislature's policymaking role: (a) the extent of its formal powers; (b) the amount of political space/discretion afforded by other power holders, mainly the Executive branch; (c) the capacity afforded by its procedures, structures and support; and (d) the goals of members and leaders of the legislature body itself. The following section explores how these factors combine to grant a specific role to the Colombian legislature in policy making.
First, the Colombia legislature performs both traditional parliamentary roles of developing legislation and scrutinizing government through both chambers of Congress -the Senate (Senado), and the Chamber of Representatives (Cámara de Representantes) -which have largely symmetrical roles and powers. Each chamber is divided into a number of commissions which deal with specific policy matters. For example, the First Commissions of Senate and Chamber of Representatives deal with 'constitutional, ethnic and peace' matters, while the Seventh Commissions of Senate and Chamber of Representatives discuss 'health, social security, housing' issues, etc. New draft laws are registered in one of the Chambers or in both (twin projects) and are allocated to a specific Committee of that Chamber for analysis and discussion. 2 Second, regarding the power of the legislature vis-à-vis the President of the Republic, the 1991 "constitution strengthened the checks and balances of the political system in an effort to endow political institutions with greater legitimacy after decades of limited participation and low representation" (Cárdenas et al. 2008: 202). This meant that "the president lost some capacity as an agenda-setter relative to the previous period, while congress and the constitutional court gained relative power" (Cárdenas et al. 2008: 202). However, the President of the Republic in Colombia continues to be extremely powerful within the Colombian system; enjoying several key powers to influence the legislation and the wider political process (including urgency message; legislative decrees; capacity to veto Congress projects; freedom to initiate laws in key policy areas, which are detailed below) (Saiegh 2010). The role of the President was further strengthened during the administration of President Álvaro Uribe Vélez (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010), who succeeded in pushing through reforms to allow him (and future Presidents) to be re-elected for a second term. Among Latin American countries, Colombia (together with Chile, Brazil and Ecuador) grants the greatest legislative powers to presidents vis-a-vis legislatures (Saiegh 2010). However, the position of the President (and by extension, the executive) versus the legislative branch should not be overstated (Saiegh 2010). It is not "imperial presidency" but rather "limited centralism" (Milanese 2011) as the President is obliged to seek compromises with parliamentarians in order to secure the passage of legislation.
Third, law making processes in Colombia are highly institutionalised. All laws, must undergo the same basic process. The first step is that the draft law is published in the official congress bulletin (Gaceta del Congreso). The process starts either in the Senate or in the Chamber of Representatives, depending on which chamber the draft was first registered. This first debate occurs in the permanent commission of the chamber after which it is voted on. If it's approved it moves on to the plenary of the chambers. Once the plenary has approved the draft is sent to the remaining chamber's permanent commission to be debated, voted and if successful, passed to the plenary. If there are differences between the approved texts in each chamber, a conciliatory draft is produced by an appointed group c onsisting ing on whether it refers to the Senate or the Chamber of Representatives. For example, the Statutory Law Project (Proyecto Ley Estatutaria, PLE) registration in Senate was PLE209/2013S and its Congress twin-project is PLE267/2013C -this is represented in this chapter in the following way: of an equal number of members of each chamber. Finally, the draft requires the President's signature to enter into law. In the case of statutory laws (and those issued under any of the extraordinary procedures), review by the Constitutional Court is required before the President's signature.
The regulations governing Congress allow for a series of extraordinary procedures to deal with specific situations. One of them is an "urgency message" whereby the President requests a higher priority be assigned to a draft law in an expedited process which should last no more than 30 days in each Chamber. The procedure means deliberation and voting on the proposed law are conducted jointly between the Commissions of the Senate and the Chamber of Representatives. Then, the draft statutory law is voted separately by each Chamber on a plenary session. A final "conciliatory" draft is produced and reviewed by the Constitutional Court to confirm that the legislation enacted is compatible with the Constitution, before being signed into by the President of the Republic. The high turnover of MPs and the weak party structures limit the institutional knowledge and technical capacity of the legislative branch in Colombia (Scartascini 2008: 47). There is very low party discipline and party leaders have only limited control of the legislative agenda (Pachón and Johnson 2016).
Finally, one of the most prominent elements of Colombia's legislature is the "personalist" nature of political candidacies (Pachón and Hoskin 2011), in which political parties serve mainly as conduits for prominent individuals, in a system favouring 'client' relations over partisan identities (Saiegh 2010;Milanese 2011;Pachón and Johnson 2016; also confirmed in interviews). Often, this involves prioritizing what Pachón and Johnson (2016) have called "distributive pork-barrel projects": obtaining resources to benefit the constituency that gets them elected (see also Milanese 2011). Saiegh (2010) highlights in addition that "legislators orientate towards satisfying narrow geographic interests". National policy makers are discouraged from making radical reforms through Congress which may affected established networks of vested interest and reforms are instead passed incrementally through executive decrees [Interview 8] or brought about through rulings of the Constitutional Court on the provision of health services (see Hawkins and Alvarez Rosete 2017).
The core of the legislature's activity is also not directed by party groups but by the Commissions of the Chamber of Representatives and the Senate, which are not under the control of party leaders but exercise significant control over the Congressional agenda (Pachón and Johnson 2016). Since the Commissions in Colombia constitute the first stage of bill approval, they "can prevent bills from ever getting to the floor. This contrasts with the situation of most legislatures in Latin America, in which committees only advise the floor with positive or negative reports" (Pachón and Johnson 2016: 73). Thus, legislators seek election to key Commissions, i.e. those which allocate resources or which play key gatekeeping functions, such as Commissions I of the Congress and the Senate in ballots held amongst parliamentarians on the first day of the legislature term (Pachón and Johnson 2016). This may mean competing for seats with members of the same party while getting support from other political groups, and again reflects the personalist, highly fragmented and decentralised nature of the Colombian legislature. Consequently, the Colombian legislature is not an arena which facilitates consensus building and constructive approaches to policymaking, with clear implications for evidence use.

coloMbIa's hIghly contested health PolIcy subsysteM
The divisions and fragmentation in Colombian society and politics are reflected in its health policy debates. The health system has suffered from a lack of fundamental consensus over its most basic organising principles and structures since its inception with the passage of Law 100 in 1993. This lack of consensus has continued throughout the almost constant process of reform which the system has undergone. Despite successive proposals for reform, high levels of political contestation have resulted in policy stasis. Deep ideological disagreements have been sustained on issues such as the financing of the system (insurance versus taxation based models); the involvement of private sector providers; and whether limits should be placed to the right to health care. This reform process reflects in part the role played by competing coalitions of actors present within the Colombian health sector, and their various attempts to shape the health system in ways amenable to their underlying interests and values. We have analysed elsewhere the interactions between three principal coalitions of actors involved in the health system reform process in the context of Colombia's antagonistic politics (Hawkins and Alvarez Rosete 2017;Álvarez and Hawkins 2018). It is possible to identify two key phases in the health system reform process. In the first phase (1993-2010), a "dominant" coalition of government technocrats, congressmen, insurance companies, the financial sector and the private health providers (including the pharmaceutical companies) emerged and was able to shape the health policy agenda. Between 2003 and 2009, under the administration of President Uribe, a set of actors began to emerge which sought to challenge the dominant coalition but were not at this juncture effectively coordinated as a coalition of actors, and so no solid agreements on a shared policy agenda between different groups were reached.
The second phase began under the current Presidency of Juan Manuel Santos in 2010 and is still ongoing. This chapter analyses the period up to 2016. A gradual coalescing of actors into distinct advocacy coalitions, increasingly coordinated and mobilised around shared beliefs and policy solutions, began to challenge those of the "dominant coalition". The emergence of "challenging" coalitions occurred in parallel with the weakening of the "dominant coalition" and the gradually weakening relationships between this coalition and successive Ministers of Health. However, as the recent passage of Statutory Law 1751 -which confirmed the principles and values of the existing health system, whilst clarifying the rights and responsibilities of patients -demonstrated, the two challenger coalitions of actors have not been powerful enough to override the hegemony of the dominant coalition and cross-coalition agreements have not been achieved. evIdence use In the fIrst Phase of health reforM (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010) Law 100 of 1993, which set up the current health system, was passed during the government of President César Gaviria (1990Gaviria ( -1994, in the context of a wider programme of the state reforms, which included the enactment of a new constitution in July 1991 (Vega-Vargas et al. 2012; González-Rosetti and Bossert 2000; Jaramillo 1998). As mentioned above, these political reforms had strengthened the power of Congress, so that it was able to impose policy initiatives on the Executive, including health reforms (González-Rosetti and Bossert 2000: 24). As González-Rosetti and Bossert comment (2000: 26): "The health reform was not part of [President Gaviria's] initial policy agenda, which focused on the social security reform. Instead, it was the concession the Executive had to make to Congress in order to have the pension reform approved." Indeed, the first version of Project Law PL155/1992S submitted to Senate by the Executive in September 1992 (and its twin draft law in the Chamber of Representatives PL204/1992C), which eventually became Law 100 later in 1993, proposed reforms of the pension system but did not include health, so Commission VII of the Senate vetoed the draft and requested the government to adopt a comprehensive approach to social security reform which also included the health system (Uribe 2009). At that time, the Minister of Health was Gustavo de Roux, who belonged to the centreleft party the Alianza Democrática M-19 and whose ideas on the health reforms were not aligned with President Gaviria's view. In November 1992, de Roux was replaced by Juan Luis Londoño, who had been deputy director of the National Planning Department (Departamento Nacional de Planeación) and who assembled a small group of national experts to take the reform task forward, supported by the group of international consultants of the Colombia Health Sector Reform Project of the Harvard School of Public Health (Bossert et al. 1998). In that same month, Londoño submitted to Congress an addition to PL155/1992S which proposed the setting up of a subsidised health scheme for the poor; but it was again vetoed by the Commission VII of the Senate, which argued again that this was only a partial reform to the health sector (Uribe 2009).
Thus, the process of negotiations on different policy options started again and a number of parties and pressure groups presented legislative projects during the first months of 1993 (Uribe 2009). These proposals were considered by a group of experts, which, under the coordination of Londoño's team, produced a new version of PL155/1992S, which was registered in Congress in April 1993. As Glassman et al. (2009: 7) state, with this proposal "[t]he administration committed to accelerating the expansion of subsidized health insurance for the poor; developing a program to support the redesign, reorganization and modernization of public hospitals and to ensure their financial sustainability; and strengthening the national immunization program." This version of PL155/1992S passed quickly through Congress between May and December, becoming Law 100 on 23 December 1993.
Key characteristics of the Colombian legislature discussed above are evident in the process of legislating the PL155/1992S into Law 100 of 1993. First, discussions within the legislature took place and a wide range of stakeholders had the opportunity to present proposals and put forward policy demands. Whilst the policy initiative came from the executive, the legislature became the central arena for these policy discussions. Indeed, the legislature managed to influence significantly the final outcome of the policy process (Uribe 2009). The vetoes to two government's versions of PL 155/1992S within Commission VII confirms Saiegh's (2010) assertion that legislatures can be active players in policy making by being blunt veto players, forcing the executive to take alternative paths.
Second, this process also shows that the different policy proposals were not submitted by the political parties with representation in the legislature, but by individual congressmen representing interest groups or social movements. This reflects Colombia's weak party system and the "personalist" nature of political candidacies explained above.
Third, evidence from scientific research was available to (and thus potentially used by) decision makers and legislators. On the executive side in particular, the consultancy team from the Harvard School of Public Health provided a continuous flow of information and knowledge at demand from Londoño and his team. According to one of our interviewees, Londoño did not want "one-off consultancy" but continuous support in designing and implementing the law and in providing answers to policy questions [Interview 8]. Research produced by academic institutions and think tanks (i.e. Fedesarrollo) was also available to other key participants of the policy process.
As an attempt to generate knowledge to support the implementation of Law 100, the government set up the Program for Supporting the Health Reform (Programa de Apoyo a la Reforma de Salud, PARS) in 1996 with the financial and technical support of the Inter-American Development Bank. The Program aimed to provide technical assistance and capacity building, to produce specialised research and strategies to transfer such knowledge to decision-makers at the Ministry of Health. More than 100 analytical studies and consultancy projects were developed until the programme finished in 2008 (MPS and Gesaworld 2008). Alongside the PARS, the strategic policy documents produced by the National Council of Economic and Social Policies (Consejo Nacional de Políticas Económicas y Sociales, CONPES) 3 on specific economic and social policy areas became a key tool to support decision making at the national level.
Despite these efforts, by 2001 the health system was facing a "severe and generalized financial crisis" leading to successive attempts to address these through reforms (Glassman et al. 2009: 7). A first wave of legislation designed to reform the health system was submitted to Congress in 2003 and 2004, but none of it was ultimately passed (Hernández 2005). The first was draft law proposed to reform Law 100 was PL180/2004S, to which other projects such as PL236/2004S and PL241/2004S were latter added for joint discussion in Congress. Supporting evidence does not appear prominently in the preambles of the draft laws, with the exception of PL180/2004s. However, the evidence is cited only in passing and without an attempt to detailed engagement. The legislative term closed in June 2004 without these draft laws having progressed, and so they were abandoned (Hernández 2005).
In 2004, a second wave of 14 draft laws was submitted to Congress, starting with PL19/2004S and followed by others such as PL31/2004S and PL33/2004S. These and the rest of the 14-strong list were accumulated to the PL52/2004S submitted by the government for parliamentary discussion. These stalled in Congress, where health insurers managed to orchestrate strong opposition, with the support of the Ministry of Finance. Furthermore, other legislative priorities such as pension reform and legislation to allow the re-election of President Uribe relegated the importance of PL52/2004S, and it was ultimately abandoned (Guzmán 2006  seek access to health service via the courts (for a more detailed discussion of tutelas and the effects of judicialisation on evidence use see Hawkins and Alvarez Rosete 2017) (Restrepo 2007;Bernal et al. 2012: 25). Law 1122 led to the creation of the Regulatory Commission for Health (Comisión de Regulación en Salud, CRES), a decision-making body at arm's length affiliated to the Ministry of Health, with the role of updating the basket of benefits through the use of high quality evidence and a strong and transparent decision-making methodology. The creation of the CRES reveals the attempt to bring more evidence into the policy making process, although this too would be subject to later reform.
Ultimately, all these initiatives to address the sustainability of the health system stalled and "by the end of the decade, the health system was in deep crisis" (Bernal et al. 2012: 25) and with seemingly little prospect of reform. Scientific research was available to policy makers in Colombia and even informed the different draft laws submitted to Congress between 1993 and 2010. However, none of these draft laws was able to pass successfully through Congress. This suggests that while evidence was important in informing policy debates, and proposed legislation, it was unable, in the context of deep politicization and embedded vested interests, to bring about effective reforms of health system and to relieve the mounting pressures it faced. and, only a few months later, in December 2012, the CRES was abolished and the Ministry of Health "re-assumed its role of resource-allocation decision-maker" (Castro 2014: 22, 131). As Dargent (2015) notes, the Santos reforms led experts and technocrats to "regain salience" at the Ministry of Health. According to Minister of Health Gaviria, the Ministry is now "a technocratic fortress", in which "decisions are now made independently of electoral politics" (Gaviria 2015). Whilst decision making of this kind is never completely apolitical, these claims speak to but a desire to introduce evidence into the decision making process more systematically and the potential for this to rationalise decision making processes.
During the second half of 2010 a new wave of reforms were introduced to Congress by different parliamentarians. Four draft statutory laws -PLE186/2010S and the accumulated draft laws PLE189/2010S, PLE131/2010C and PLE198/2010S -which aimed to define "the essential core of the right to health" were discussed in Parliament. Research evidence was provided in the preambles of 3 of the 4 draft laws submitted to Congress. PLE186/2010S analyses in detail the dramatic increase of health care costs and its causes, while PLE189/2010S included an extensive commentary on the vision and recommendations of the World Health Report 2008 on primary care (WHR 2008). PLE198/2010S provided figures on equity and access to health services based upon official statistics, for example, and includes references to specialised literature to back up the proposals suggested. All these draft laws however did not complete the process in a single legislature and hence had to be abandoned.
A new window of opportunity for policy change opened up in the summer of 2012 with the submission to Congress of four draft statutory laws (PLE48/2012S; PLE59/2012C; PLE105/2012S; and PLE112/2012S) and one ordinary draft law (PL51/2012S, which was consequently accumulated to PL210/2013S which came later). Of the four statutory laws, PLE48/2012S was informed by good quality research obtained from international comparisons, published studies and interviews with experts. PLE105/2012S and PLE112/2012S also referenced publications and official data, while PLE59/2012C did not mention any specific research.
With the impetus brought by the newly appointed Health Minister Gaviria, the government sought to take the initiative on the reform of the health system and develop both a Statutory Law and an Ordinary Law.

conclusIon
The analysis of the more than 20-year long process of reforming Law 100 shows that evidence can, and indeed did, inform health policies in a highly contested and fragmented political setting. The analysis of draft laws designed to reform the health system shows that policy relevant evidence was available to actors involved in reforming the health system and was used to inform a number of key draft laws submitted to parliament. We identified high levels of contestation and fragmentation at different levels of Colombian society: the social, political and policy levels, which provide the institutional context in which policy problems emerge and policy actors seek to address them. In this context, the availability of policy relevant evidence offers a potential means of circumventing and overcoming, political fragmentation and contestation. However, the deep seated nature of the vested interests in the Colombian health system and the health systems models which they favoured, meant that reform proposals were often stymied. Whilst it is not possible to depoliticize or solve policy dilemmas through recourse to evidence alone, it is possible, at times, to use evidence as a means of generating consensus or providing the impetus towards compromise. To overcome the endemic problems of weak political and legislative structures, and engrained political cleavages, new bodies were formed which were tasked with the collection, interpretation and deployment of policy relevant evidence.
However, despite the institutionalised mechanisms of evidence generation and synthesis, the translation of research into policy and legislation remained limited and piecemeal. Many draft laws provided only minimal data on the extent of the problems facing the health system in Colombia, and their reform proposals were mostly based upon general commentaries of the functioning and challenges of the system (i.e. PL229/2010S; PLE59/2012C; PL51/2012S; PL233/2013S; etc.). A small number of other draft laws, which quoted academic studies to back up their analysis of the extent of the problem, tended however to refer to a biased selection of studies (that is, research which would fit their ideological stances; i.e. PLE105/2012S; PLE112/2012S). However, those draft laws submitted by the government, especially during the last phase of the reform, included extended empirical and analytical sections. For example, the PLE209/2013S and PL210/2013S provide a deep and wide analysis of the problems of the Colombian health system.
Reflecting the role of the legislature in Colombia's highly contested political system and health policy subsystem, evidence cited in draft laws was unable to forge consensus amongst relevant policy actors over the direction of the health system reforms. The deep confrontations within the Colombian legislature did not facilitate political agreements nor play a constructive role in health policymaking. Scientific research was available and at the disposal of legislators, but it was unable to provide the common ground on which to overcome embedded policy positions and form the basis of compromise over the direction of health systems reforms.