3 - The Unrealized Potential of Presidential Dominance in Colombia

5/25/2018 3 - The Unrealized Potential of Presidential Dominance in Colombia - slidepdf...

http:///reader/full/3-the-unrealized-potential-of-presidential-dominance-in-co

Cambridge ooks Onlinehttp://ebooks.cambridge.org/

Presidentialism and Democracy in Latin America

Edited by Scott Mainwaring, Matthew Soberg Shugart

Book DOI: http://dx.doi.org/10.1017/CBO9781139174800

Online ISBN: 9781139174800

Hardback ISBN: 9780521572668

Paperback ISBN: 9780521576147

Chapter

3 - The Unrealized Potential of Presidential Dominance in Colombia pp.

110-159

Chapter DOI: http://dx.doi.org/10.1017/CBO9781139174800.004

Cambridge University Press



THE UNREALIZEDPOTENTIAL OFPRESIDENTIALDOMINANCE IN COLOMBIARonald P. Archer and Matthew SobergShugart3 .1 . INTRODUCTION

On July 4, 1991, Colombians ceased tolive unde r wh at had been theoldest continuously functioning constitution in Latin America. A newdocument, drafted and ratified by the specially elected Constituent Assem-bly, replaced a constitution that, while frequently amended, had served since1886. Inorder tounderstand why C olombians un dertook the processofconstitutional revision, one needs to consider the various ways in which theold constitution had contributed toacrisis tha t m ost C olombian politicalobservers believed was confronting the country by the late 1980s.Asig-nificant aspect of the C ons titue nt Assem bly's task was to remedy a perceivedimbalance between anoverly powerful presid ent and aCongress th at ap-peared incapable of addressing the nation's problems.

In this chapter we argue that Colombian presidents hold great consti-tutional powers over legislation yet frequently appear unable to accomplishpolicy agendas that are nominally endorsed by their own parties. Despitemost presidents' having held copartisan legislative majorities, presidentshave low partisan powers because parties are internally fragmented. Twoinstitutional features contribute tolow partisan pow ers: First, parties lackcontrol over the use of their pa rty labels (i.e., candidates do not need to haveparty approval torun und er t he pa rty nam e); second, the electoral systemfosters intraparty competition.

110

Downloaded from Cambridge Books Online by IP 201.231.200.165 on Mon Sep 17 19:03:41 BST 2012.http://dx.doi.org/10.1017/CBO9781139174800.004

Cambridge Books Online Cambridge University Press 2012



Cambridge Books Online Cambridge University Press, 2012

PRE SIDENTIAL DOMINANCE IN COLOMBIA 111Evaluations ofpresidentialism in Colom bia have tended to overlook

the real limits to the powers ofthe Colom bian executive. Ou r argum entis that the partisan powers of the Colombian president have decreased sincethe mid-1970s as a result of increased intraparty com petition . T hisin-creased competition in turn has been fueled inpart bythe g row ing diffi-culty of reconciling em erging urban de man ds w ith those of moretraditional sectors. The result isthat presidents m ust relyon adhoc dealswith power brokers who demand patronage in exchange for supportingpolicies desired by thepresident. So,despite vast constitutiona l powerssince an im porta nt constitutio nal reform in 1968, real ability to effectchange hasbeen lower tha n thedesignation ' 'potentially d om ina nt" (seeChapter 1)would imply. W econceive ofthe p reside ntial powers definedin theconstitution in terms ofprincipa l-agent theory,1 as a delegationfrom Congress to theexecutive inorder toadvance the goals ofcongres-sional majorities. W e argue that such aninterpretation isvalid despitethe separate election and fixed terms ofpresidentialism because, uniqueto Colombian presidentialism, legislative majorities determine the alloca-tion of constitutional powers through control over the amendment process.Thus , tounderstand Colom bian presidentialismit iscrucial tounderstandthe interests ofCongress andhow theformal Colom bian political insti-tutions shape those interests.

The C olombian Congress is structured in a way that makesitinherentlydifficult to confront national policy issues. Instead, members' incentives areto cultivate particularistic exchanges based on patronage. The resulting in-ability ofthe Colom bian political system togenerate stable m ajorities hasmadeitdifficult to confront Colo mb ia's profound social prob lem s, especiallythe country's infamous rural and urban violence. Members of the majorityLiberal Party face a collective action problem.2 Individual memberscan1 A core component of the theory of the firmineconomics holds th atprincipals- usuallythose with power and/or wealthcan offer binding contracts toagents,who then performspecified services (Williamson 1975, 1985; for extensions to political science, see Kiewietand McCubbins 1991: 239; Moe 1984). Certain features hold in virtually all principal-agent situations: (1) The principal retains mechanisms to oversee and verify that the agentis performing the specified service; (2) delegation entailsacertain degree of agency lossor slack, where the agents can hide information or use the delegated powers against theprincipal, among other things,2 First identified by Olson (1965),a collective action problem arises whenagroup of in-dividuals attemptstocoordinatetoprovide collective benefits, or public goods, for thewelfare of the whole group. Since providing the collective good iscostly and, once thegood is provided, each individual can benefit from the public good given its public nature,individuals are likely to "free ride" on the work of others- doing nothing to provide thegood, while still receiving the benefit.






112 R O N A L D P . A R C H E R M A T T H E W S H U G A R Tunderstand the need for their party to back fundamental reforms that wouldaddress the interests of a growing urban electorate, because (as we shallargue later) this electorate is crucial to the party's ability to maintain controlover the executive. Yet most members are tied by the electoral system togenerally rural clientele networks that demand access to patronage. Mem-bers of Congress must participate in this scramble for patronage, or theirpolitical careers are doomed. Thus, the dominant incentive for members isto defect from reformist coalitions or to join them only after they have beenwatered d ow n or laden wit h sufficient p atron age resources. W e should cau-tion here that we do not mean to imply that Colombia's many social prob-lems can be blamed on its political institutions, but we do claim that theseinstitutions and the incentives they establish have seriously inhibited effortsto cope with these problems.

Seriously needed reforms become bogged down as they wend their waythrough the Colombian policy-making process. The irony of the process isthat the reforms are needed not just from a normative perspective, but froma political one. A series of popular movements and paroscivicos (civicstrikes) have articulated demands that are frustrated by the clientelisticpolitical system (Coral Quintero 1988; Gallon Giraldo 1989; Pecaut 1988;Santamana and Silva 1984). Colombia's growing urban population de-mands specific reforms such as building urban housing and infrastructure,creating a more professional civil service to replace citizens' need for po-litical connections in order to receive many government services, reformingthe judiciary so that it can deal with drug traffickers and official malfea-sance, and cleaning up contamination of the country's water and other nat-ural resources. Because the political system has been unable to deliver onpresidential promises to initiate reforms of these types, urban voter ab-stention is high, and by the late 1980s, the country's political stabilitycould no longer be taken for granted as conflict became more and moreviolent.

The rest of this chapter is organized as follows. In Section 3.2, weoutline the series of reform efforts by Colombian presidents that have endedin defeat and we summarize our overall argument. In Section 3.3, we de-scribe the constitutional powers of the Colombian presidency. In Section3.4, we identify the electoral system and how it contributes to low partisanpowers for the presidency. Section 3.5 then discusses how the party systemaffects presidentialcongressional relations and why Colombia is character-ized by a relatively reformist executive and a more conservative legislature.Section 3.6 considers whether the new constitution provides a solution tothe dilemma of reform in Colombia. Section 3.7 concludes.






PRESIDENTIAL DOMINANCE IN COLOMBIA 1133 . 2 . THE THWARTED REFORMEFFORTS OF COLOMBIANPRESIDENTS, 1974-90Each president since the endofthe N ationa l Front po wer-sharing arrange-men t3 has proposedapackage of reforms inten ded to"open" and ' 'mo dern-ize" the political system. These reform packages maybeseenin large partas attempts bypresidents totranscend therural, clientelistic basisofrep-resentation in theCongress and theatte nd an t "inefficiency" of a publicadministration that is based heavily on patronage. Each president found hisreform package mostly stymied in theCongress andthus a t tempted toimplement some programs bybypassing thelegislature altog ether. Asweshall seeinthis chap ter, the substantial decree powers granted tothe pres-ident in the Colombian constitution generally have proved ineffective when-ever a presid ent tried t o use them to step outside the rath er narrow consensusestablished by Colombian party bosses with their tiesinclientele netwo rks.Colombian presidents have been able topush policy outcomes in a more"progressive" direction than Congress likely would have done on its ownas Chapter 1 suggested proactive presidencies could do but the end resulthas usually beenfarless tha n wh at presi dents have tried toaccomplish.President Alfonso Lopez Michelsen (1974-78) launched a programcalled To Close the Ga p, which was intended to redress Colombia's grow ingincome inequality and to restructure thepublic adm inistration (Gomez1978). Mostofthe p rog ram was defeated inCongress (Dix 198 7:19 0-93 ).His attempt to establish two redistributive institutions, the DRI (DesarrolloRural Integrado) and PA N (Plan de Alim entacion y N utrici on) were quicklypassed byCongress andalmost asswiftly cann ibalize d forpatronagere-sources (Pecaut 1988:25455). Infact, income dis tributi on actually wors-ened under Lopez (Pecaut 1988:3045), and spending on health, education,and welfare remained comparatively low (Dix 1987:200201).

Frustrated by thedefeat orwatering down of so many proposalsinCongress, Lopez,inwha t has becomeafamiliar them efor Colombian pres-idents, attem pted an end run around the legislature via a constituen t assem-3 The N ational Front was designed to reconcile conflicts between the country's two dom i-nant political parties, Liberal and Conservative, which had led to civil war and thena

four-year military dictatorship. The major provisions of the Front were: alternation of thepresidency between thetwo parties every four years un til 1974; guaranteed parityofrepresentation in allnational and subnational legislative bodies, courts, and in the bu-reaucracy; and, until 1968, the requirement oftwo-thirds majorities forthe passage oflegislation. For details, consult Berry, Hellman, and Solaun (1980), Dix (1967), and Har-tlyn (1988).






114 R O N A L D P . A R C H ER M A T T H E W S H U G A R Tbly that would pass reforms in the areas of departmental administration andthe judiciary. The effort was nullified by the Supreme Court4 as an uncon-stitutional usurpation of Congress's powers of amendment (Arboleda 1982:1131). In one of Lopez's few successes, the Supreme Court upheld hiscreation of a more progressive income tax program by decree (Pecaut 1988:2 50 -5 2 ). Thi s example of reform via decree is mo re exceptional tha n typ ical,however, as we discuss later on in this chapter.

President Julio Cesar Turbay Ayala (197882), although not normallyperceived as a reformist, picked up many of the failed Lopez reform initia-tives and sent his own package to Congress. Turbay, despite his reputationas a master at crafting clientele coalitions, fared no better than Lopez. Hisreform package was even more ambitious and extended to reform of Con-gress and the nation's judicial, banking, and economic planning institu-t ions. Again, the reform project was stymied by the Congress, and an endrun effort to circumvent Congress was declared unconstitutional (Arboleda1982:33-60) .

By 1982 the clamor from public opinion and the mass media for therestructuring of local, departmental, and national governmentsperceivedas unresponsive to voters allowed the minority Conservative Party to wina contested presidential election for only the second time since 1926. 5 Ledby Belisario Betancur (198286), the new government combined the Lopezand Turbay reform p rogram s wi th an amb itious peace plan to end the armedinsurgencies that were plaguing the countryside. The Betancur program,therefore, was even broader in scope, calling for (1) a redefinition of the roleof opposition parties; (2) campaign finance reform, the institutionalizationof the role of political p arties, the r igh t to inform ation, and electoral reform;

4 The Colombian judiciary has one ofthemost politically insulated appointm ent proceduresin the world. Since 1958, the president and Congress have not had their previous role innominating and confirming justices. Instead, vacancies are filled by the remaining mem-bers themselves. Justices had life tenure until the new constitution limited them to asingle eight-year term. The same procedures that are used for the Supreme Court are alsoused for the country's highest administrative court, the Council of State. The new con-stitution establishes the Superior Council of the Judicature, some of whose members areappointed by the Court and Council of State themselves and some of whosemembers arenominated by the president and confirmed by Congress. This judicature council will sendto the two highest courts slates of nominees from which they will fill their own vacancies.Finally, the new constitution establishes the Constitutional Court, which will be farlessinsulated politically: It will be composed of members elected for nonrenewable eight-yearterms by the Senate from slates of nominees submitted by the president and the other twohigh courts.

5 Bycontested electionwe mean (1) no boycott by one of the major parties, as in 1949 whenthe Liberals abstained, and (2) no restrictions on candidate entry, as under the NationalFront.






P R E S I D E N T I A L D O M I N A N C E I N C O L O M B I A 115(3) a restructuring of the state; (4) a redefinition of executive emergencypowers; and (5) the creation of new means of local and departmental popularparticipation in government, such as the direct election of mayors, publicservice user groups, and district councils (Santamana and Silva 1984:73187). By 1986, however, virtually all of these initiatives had fallen victimto congressional recalcitrance and the collapse of the peace talks with theguerrillas. Betancur also tried to implement a major tax reform by emer-gency decree, but the Supreme Court ruled it unconstitutional.

The successor to Betancur, Liberal Virgilio Barco Vargas (198690),was to fare little better in pursuing his reform agenda. Picking up thegauntlet from three former administrations, Barco attempted to craft a broad"coalition-of-the-whole" with the major leaders of all the parties and partyfactions around his proposal to repackage the Lopez, Turbay, and Betancurreform programs and get them through Congress not as discrete pieces oflegislation, but as part of a large package of constitutional amendments.When the original agreement among the parties collapsed, so too did theconstitution al reform package, now containing m ore than a hund red am end-ments (Semana, May 31, 1988:3644). Beyond the previous reform initia-tives, Barco expanded the reform process to include treatment of churchstate issues, property rights, access to education, the establishment of aprocedure for holding national plebiscites, and many other reforms designedto open up a democracy still operating u nder m ultip le anachronistic restric-tions from the National Front era (Cepeda 1987). In the end, Barco, likehis predecessors, attempted to circumvent Congress, proposing the creationof an essentially unconstitutional constituent assembly (const tuyente).By198 9 the pressure for change had grow n so great from nearly all of the majorforces in Colombian society that the Supreme Court closed its eyes to theillegality and approved the creation of the assembly.In 19 91 , under the ad min istration of President Cesar Gaviria Trujillo(1990-94), the fifth president to attempt significant post-National Frontreform, the pent-up reform initiatives were finally written into a new con-stitution with an impotent Congress on the sidelines, criticizing the entireprocess. The Constituent Assembly, elected in December 1990, includedrecently demobilized guerrillas such as the M-19, who won over a quarterof the votes cast. The Assembly even went so far as to dissolve the Congresselected in March 1990 and call congressional elections two and a half yearsearly.

The country's new constitution thus is the culmination of nearly twodecades of political reform efforts by presidents and leaders of both majorparties. Party leaders have recognized the political necessity of respondingto the mobilized constituency for change, but until President Barco's decree






116 RONALD P . A R CH E R M A T T H E W S H U G A R Twas uph eld by the Suprem e Cou rt, each at te m pt had fallen prey to a congressmore interested in defending a status quo on which they depended for accessto patronage. The new constitution makes some important changes thatappear to be weakening the old clientele netw orks. Some encouraging trendsinclude thepost-Constituyente passage of legislation that gives parties some-what enhanced control over the use of party labels and anational senatedistrict that allows new parties or factions of the traditional parties to attractsupport that transcends the old electoral districts, where traditional clien-telistic brokers effectively shut them out before. Thus, the interests repre-sented in Congress are becoming more diverse; yet the increase in intrapartycompetition continues apace, and in suchacontex t t he reduced po wers ofthe presidency may make governability more difficult. Thus, the early post-reform verdict must be a mixed one.

CONGRESS AND THE PROBLEM OFREFORM THE ARGUM ENT IN BRIEFThere is acertain irony inthe just-reviewed attem pts toreform pub licadministration and the Congress: In nearly every case presidents have spon-sored reforms that would have weakened their own powers, especially byrestraining their emergency powers and by reducing the patronage resourcesthat have long been the main "carrot" offered to members of Congress inexchange for their votes. This irony forms the core of the argument that wemake in this chapter, which is that the seemingly "powerful" Colombianpresidency has been crafted in suchaway as to serve prima rily the clien-telistic interests that are represented in Congress. In other words, the po-tential dominance of the Colombian presidency between 1968 and 1991often remained more potential than real, because the Congress was able toprevent presidents from effectively using their powers to move beyond whatlower-ranking party leaders would tolerate.

As noted earlier, prior to the convoking of the Constituent Assemblythrough astate-of-siege decree in 19 90 , the only means for reforming theconstitution was for the Congress to pass, in two consecutive regular ses-sions, alegislative act by ab solute m ajorities of both houses.6 Thus, the6 The new cons titution keeps this procedure, which is unusually easy compared with otherpresidential constitutions. It is possible to reform the new constitution by means of eithera plebiscite or a constituent assembly, although neither path can be taken without priorcongressional authorization. In addition, amendments to certain parts of the constitutioncan be overturned by a plebiscite initiated by 5% of registered voters after Congress hasenacted them. See Articles 374-80.






P R E S I D E N T I A L D O M I N A N C E I N C O L O M B I A 117constitutional allocation of powers among the branches was a matter underthecontrolo f theCongressalone\ neither a referendum nor the consent of statesor provinces (known as departments in Colombia), nor even a supermajorityof the legislature, was needed. Only in 199091 was the congressional sov-ereignty over the constitution abrogated through the extraordinary proce-dure of convoking a constituent assembly.

Therefore, we need to understand why legislators have resisted reformsthat would enable them rather than a decree-monger ing president toclaim credit for efforts to respond to mobilized popular clamor for reforms.And to understand this problem requires an understanding of what theinterests of members of Congress are. In this chapter we demonstrate thatthese interests are primarily grounded in the provision of patronage to cli-ents that is, to a narrow subset of the voters in the constituency fromwhich they are elected7 rather than being accountable to the larger elec-torate on the basis of policy programs. The delegation of extraordinarypowers to the president, when viewed from this perspective, is merely ameans of furthering these clientelistic interests by freeing members fromthe burdens of making policy. If, because of the way that they are electedfrom narrow clientele networks, members are not being held accountablefor national policy anyway, then they have little incentive to invest timeand effort in building and sustaining policy coalitions. Instead they canfocus on extracting resources to reward supporters with jobs and other pri-vate goods. They will tend to involve themselves in policy only when thepolicy proposal under consideration furthers their primary interest in con-verting public funds into private payoffs. Indeed, the Colombian Congresshas from time to time enacted significant policy reform, but the successfulexamples underscore our point: These reforms, such as the direct electionof mayors since 1988 and administrative decentralization, have actuallygiven members new opportunities to build their own careers via the pro-vision of particularism. On the other hand, most of the reforms containedin the various presidential policy packages sketched earlier would have cur-tailed the ability of members of Congress to use national resources to cul-tivate the patronage networks on which their own careers depend.Therefore, the political system rarely has responded to demands for po-litical reform.

7 As noted in Section 3-4, members of the lower house (and before 1991 , of the Senate too)are elected from multiseat districts that are coterminous with the country's administrativeunits, knows as departments. Since 1991 the Senate has been elected from a nationwidedistrict.






118 RONALD P. A R C HE R M A T T H E W S H U G A R T3 . 3 . P O W E R S OFT H EC O L O M B I A N P R E S I D E N TThe concentration of formal power in the central government and the ex-ecutive has led some scholars to characterize the Colombian president asa"Demi-God" (Cepeda 1985). The powers of the president were expandedby reforms dur ing the N ation al F ront coalition governme nts (195874) andmost especially wi th the constitutional reform of 196 8. W it h th e prom ul-gation of the 1991 constitution, new constraints have been placed on thediscretion of the executive in most policy areas.

Speaking of presidential powers from the National Front up throughthe 1980s, Vazquez Carrizosa (1986:15) said that "the President of Colom-bia isalmost the totality of the s tat e" and t hat his "decision powers arealmost abso lute." W hil e we argue that this statem ent is a clear exaggeration,it is indeed true that the presidency is the source of nearly all substantivelegislation, much of which isnot even deba ted inCongress because it isenacted under emergency powers. Indeed, it is the regular use of what aresupposed to be "exceptional" powers that has given Colombian presidentstheir ability todo m ina te the legislative process. Yet emergency powersamount to less than meets the eye whenitcomes to attem pts by presidentsto carry out structural reforms. Before developing this argument further,itis necessary toconsider the formal powers of the presid ent, w hich can begrouped into four different categories that, while interrelated, are analyti-cally distinct. These are (1) powers of appointment, (2) legislative powers,(3) fiscal and planning powers, and (4) extraordinary powers. 8

POWERS OF PPOINTMENTThe Colom bian presiden t freely names and m ay dismiss all ministers, othertop- and medium-level employees ofthe executive bran ch, amb assadors,directors of top state "decentralized institutions" (such as the state petro-leum company), and the directors and boards of various financial institutionsthat are intervened in or controlled by the state. 9 The new constitution has8 This does not exhaust all possible presidential powers. For instance, the president is com-mander of all forces of public order, which gives him some discretion in opening or closing

spaces for public protest. He also has, as head of state, powers in foreign policy, such asnegotiating treaties. Under A rticle 224 of the 1991 constitution, he may even put treatiesrelated to international trade in to effect provisionally, subject to ex post veto by Congress.9 The president continues toappoint all directors of the Central Bank, although thesemembers now have staggered terms. The staggering of terms isnot very significant,however, given that the terms are of the same length as the president's, and thus everypresident has the opportunity to change every member of the board of directors during






PRESIDENTIAL DOMINANCE IN COLOMBIA 119

somewhat diminished thepresident's autho rity over thecabinet. Hestillmay appoint and dismiss ministers without approval ofthe Congress, butCongress also may remove ministers throughamo tion of censure. Th us,apresident's ability to maintain acabinet to his liking has been weakened,although it still falls tothe president to decide upon areplacement foracensured minister.10

Before 1991 the president's powers ofapp oint m ent and dismissal ex-tended tothe governorships of all of Colom bia's departm ents. Since 199 1,however, these positions have become elective. Thus, an enormous patronageresource of the president has been eliminated under the new constitution, 11whileanim porta nt avenueofparty and po litical career buil din g has beenopenedup toelectoral com petition .

LEGISLATIVE POWERSAs noted inChapter 1,apresi den t's powers over legislation fall into twobroad categories: reactive (such asaveto) and p roactiv e (such as decree andemergency powers).

Reactive PowersThe veto power of the Colombian president isone of the m ajor areasin

which the 1991 con stitution has reduced presidential powers compared w iththoseofthe former constitu tion. U nde r the pre-19 91 co nstitutional order,on certain kinds oflegislation the pre siden t's veto could beoverriddenbyonlyatwo -thirds vote of both houses of Congress. The list (A rt. 88 /188 6)1 2of such types of legislation consisted of legal codes, the organic rules for thenational budget, changesinterrito rial subd ivisions of the co untry , and th e

his term . The board has seven members: the d irector, chosen by the board and a memberof that board; the ministeroffinance(hacienda);and five individuals serving four-year,renewable terms. At least two of the six nonministerial members must be replaced everyfour years, but the president has the prerogative of determining which members thesewill be (Art. 372 of the 1991 constitution).10 For a discussion of censure in comparative perspective, see Shugart and Carey (1992).11 Until 1988, mayors were appointed by governors, thus constituting asort of indirectpatronage resource for the president. Since 1988, however, mayors have been popularlyelected.12 We shall use the following format throughout this chapter when citing constitutionalarticles. The article number will be followed, where necessary, by the paragraph numberafter a colon. The constitution in question, either 1886 or 1991, will be indicated aftera slash. Any reference to the 1886 constitution refers to the version in effect in the 1980s(Giraldo M. and Gomez Velasquez 1983), unless otherwise indicated. References to the1991 constitution are based on Vidal Perdomo and Sachica (1991).






120 R O N A L D P . A R C H E R M A T T H E W S H U G A R TNational Development Plan.13 All other types of legislation under the for-mer constitution could be overridden by a majority of all members of eachhouse. Thus, aside from the policy areas just listed, the veto is a relativelyweak one if the interests of a majority party (or cross-party coalition) on agiven bill is sufficiently compelling to overcome the chronic low attendancethat plagues the Colombian Congress. The new constitution provides forsuch a formally weak veto in all areas of legislation.

The Colombian president under both constitutions has been able toindicate in his veto ofabill that he objects to only part of the bil l. H owever,while the president may request that Congress reconsider parts of a bill, hedoes not have the power of partial promulgation; that is, he cannot enactinto law those parts of the bill to which he does not object. Upon thepresident's partial veto the entire bill is returned to the chamber of origin,which can discuss only the objections raised by the president. The Congresscan then either accept or reject the president's partial veto. Should it acceptthe president's veto, it then sends the bill in its changed form to the pres-ident's desk for promulgation. Should it muster the votes required to over-ride a veto, it then returns th e bill to th e president in its original form. T hepower of partial veto, but without partial promulgation, gives the presidentsomewhat greater leverage in dealing with the Congress. The president canbe certain that the parts of the bill to which he does not object will not beeroded by new amendments when Congress reconsiders the bill, becauseonly the parts that the executive wants changed can be debated again. How-ever, this provision gives the president less leverage than if the parts of thebill he preferred could be promulgated before the Congress reconsidered thevetoed parts.

In both Colombian constitutions, the president has been granted theexclusive autho rity to in itiate legislation in certain policy areas. Like a veto,this power is essentially reactive, in that it allows the president to preventthe Congress from changing the status quo unless the president submits aproposal to change it. The following policy areas have been reserved forpresidential initiatives in both constitutions: laws on the structure of the13 The National Development Plan (NDP) was, according to the 1968 amendments, toestablish a general statement of long-term, medium-term, and immediate goals and abudget for all national public entities to carry out these goals. Article 341/1991 - a

virtual copy of the corresponding article from the previous constitution - permits thepresident to enable the NDP by decree if the Congress has not come to agreement on itwithin three months of its submission. Such tremendous executive power may explainwhy the Congress never followed up by passing the implementing legislation to put theprovisions of the NDP into effect. Thus, in the absence of enacting legislation, thiselement of the constitution itself remains no more than a statement of the principle thatthere should be an NDP.






P R E S I D E N T I A L D O M I N A N C E I N C O L O M B I A 121executive ministries; laws establishing salaries for public officials; and gen-eral norms for regulating foreign exchange, external trade, the national debt,and tariffs (Arts. 79/1886 and 154/1991). In the 1886 constitution, but notin the 1991 constitution, the president also had the exclusive right to ini-tiate legislation on the organic rules of the national budget and publicworks. Note that these were areas in which the 1886 constitution also gavethe president a veto with a legislative two-thirds override. Thus, in thesetwo policy domains, the president could be almost certain that his ownpreferences would be translated into law, because (1) Congress could con-sider no proposals other than those of the president, and (2) if a majority ofCongress mem bers (but less than t he difficult-to-obtain tw o-third s majority)amended the bill to such a degree that the president no longer consideredthe bill an improvement over the status quo, he could always veto it.

The 1991 constitution has granted the president the exclusive right tointroduce legislation in two new areas in which Congress previously hadthe power of initiative.14 However, the more important point about the1991 constitution is that the president's powers to shape the legislativeoutcome have been reduced in two ways. First, important areas such as theorganic rules for the budget and public works are no longer on the list ofexclusive presidentia l initi ative . Second, as noted , in all issue areas the pre s-iden t's veto is now w eak. In those areas wh ere the presid ent has the exclusiveinitiative, he can still prevent changes to the status quo simply by choosingnot to initiate a bill, but once he initiates it, Congress is free to amend it,and a determined majority could prevail.Proactive Powers

It is in the area of proactive powers that Colombian presidents haveearned their reputation as the chief legislators as a force for changing thestatus quo, not merely preserving it. The decree powers of the Colombianpresident are extensive, especially if we include under this category emer-gency powers. These powers will, however, be discussed separately later.Here we speak only of the president's ordinary decree powers that is, theability to enact decrees even in the absence of a declared state of exception.

An important area of presidential decree authority is the pro tempore

14 These are grants of authority to negotiate contracts and loans and to establish the revenuesand expenses of the administration. In addition, the president may exclusively initiatelaws concerning the functioning of the Central Bank; however, previously the presidenthad unrestricted decree power in this area. Thus, the new constitution, by giving theCongress any right to legislate on matters concerning the Central Bank, has reducedpresidential powers. We discuss the president's fiscal and planning powers in an upcom-ing section.






122 R O N A L D P . A R C H E R M A T T H E W S H U G A R Tauthority (Arts. 76:12/1886 and 150:10/1991), which consists of specialpow ers, delegated by Congress, to issue decree-laws du rin g specified periodsand on subjects defined in the legislationitself.Such an autho rity has existedsince the 1886 constitution went into effect, although the precise ways inwhich it could be used have been redefined over time. For instance, since a1936 constitution al am end m ent, th e Colom bian state has been charged w ithintervening in the functioning of the economy to ' 'rationalize" productionand the distribution of resources. A 1945 amendment clarified that thiscould not be accomplished through delegated decree authority, but onlyunder specific legislation passed by Congress. In 1968, however, anotheramendment explicitly allowed the president to use decree-laws to intervenein the economy, as long as these decrees were under the authority of a lawpassed under Article 76:12/1886 (Findley 1980).

Although the purpose of delegated decree authority is to allow Con-gress, as principal, to empower the executive, as an agent of Congress, toundertake measures to further the interests of the Congress, in practice anyprincipalagent relationship is bound to entail some degree of agency slackor agency loss (Kiewiet and M cCub bins 1991). Indeed, there are numerou sexamples, especially after 1968, of presidents' pushing to the limits theirdelegated authority under Article 76:12/1886. One of these examples,drawn from a more extensive study by Findley (1980), will suffice to dem-onstrate the problem. Colombia has long suffered from a shortage of urbanhousing, in part because of a shortage of long-term credit. President CarlosLleras Restrepo in 1968 proposed a bill to create a National Savings Fundinto which employers would be required to deposit, in advance, funds thatwould be equivalent to the severance pay rights (one month's pay for everyyear worked ) of all current worke rs. Congress, und er pressure from businessinterests, rejected the bill.However, in spite of congressional rejection of this measure, the pres-ident implemented through decree a limited form of the policy for govern-ment employees only. Lleras cited Law 65 of 1967, which delegated powersunder Article 76:12/1886 to reorganize the executive branch. On the basisof a clause in the law tha t authorize d t he president t o ''fix . . . the regim eof social services" for government employees, the Supreme Court ruled thedecree valid.

Thus , in this example, the president was able to enact a scaled-downversion of a policy that had been explicitly rejected by Congress, because ofan earlier act of delegation by Congress that could be interpreted as givingthe president authority to legislate in the area. Article 76:12/1886 did notgive the Congress the authority to pass legislation that would contradict adecree issued under the authority of that article, so the policy stood. Thus,






P R E S I D E N T I A L D O M I N A N C E I N C O L O M B I A 123under the pre-1991 constitutional order, the president could enact decreesthat established a new status quo that Congress could not rescind or modify,short of a constitutional amendment.The 1991Constituyente put several important restraints on this power.Theproactiveness of the pro tempore decree power has been reduced in thatCongress now has the explicit authority to repeal or amend a decree issuedpursuant to delegated faculties (Art. 150:10/1991). The same article alsostates that Congress may delegate such faculties for a maximum period ofsix months previously there was no stated limit and only in areas ex-plicitly requested by the government. A bill establishing an area to belegislated by decree must pass with a majority of the whole membership ofeach house, rather than by just a majority ofaquorum. Finally, pro temporepowers may not be used to decree codes, statutes, organic laws, or taxes.The Constituent Assembly thereby attempted to restrict Congress's pro-pensity to delegate excessively, forcing it alone to come to terms with themost important legislative matters.

In addition to the power to enact legislative decrees, the president alsohas the authority to regulate bills passed by Congress. By means of a reg-ulatory decree, the executive may define the precise areas to be affected bythe legislation passed by Congress. An excellent example of the use of reg-ulatory decrees can be found in the process of adm inistra tive decentraliz ation(for a general discussion, see Dugas, Ocampo, Orjuela, and Ruiz 1992). InArticle 13 of Law 12 of 198 6, Congress gave the executive branch th e power(for a period of one year from the pro m ulg atio n of the law) to write a n um berof qualifications left unstated at the very heart of this particular piece oflegislation. Law 12 of 1986 involved the devolution of proceeds of thenational sales tax (VAT) from the central government to municipalities. Atthe same time, the national government also turned over to the municipal-ities various functions previously under the control of the central or de-partmental governments and closed down now-superseded national ordepartmental public service institutions. However, Article 13 of Law 12left to the executive the right to determine what functions would be turnedover to the municipalities and which institutions would be closed.

In Decree 77 of 1987, for instance, the president mandated a curtail-ment of the budgets and functions of various governmental institutions. Atthe same time, many functions were now expressly turned over to the mu-nicipalities, including providing potable water, basic sewerage facilities, andmunicipal slaughterhouses and marketplaces. A review of the original lawand the subsequent enabling decrees illustrates some interesting and tellingpoints. First, the original law was only 12 pages long, while the variousexecutive decrees ran on for 72 pages. Second and m ore intere sting from the






124 RONALD P . ARCHER MATT HEW SH UG AR T

perspective of this chapter, the Congress left in thehandsof the executivebranchthe tough decisions about which institutions wouldbeclosed, w hichwouldbe cut back, and which would see their budgets grow. In thispar-ticular case, while the various clientele networks composing Congress allsupported thedevolution oftax mon ies from thecentral government to themunicipalities, they could notagreeon thedetailsof the transition process.Rather than chance thelossof the greater gain, they madeuse of the insti-tutional structures establishedby theconstitution topasson this politicallydifficult taskto the executive branch.15

In recent yearsthe regulatory powersofthe presid ent have increasedformany of the same reasons that the president's legislative powers have in-creased:thelackoftechnical capacityandpolitical willon thepartofCon-gress toconfront ma ny ofthe difficult political decisions necessarytoregulate existing legislation. Reasonsfor this lack'ofwillon thepartoft heCongress willbeaddressed in Section3.4.

FISC L ND PL NNING POWERSO neofthe m ost im porta nt sources of presidential power inColombiais theextensive controltheexecutivehasoverthenational budg etand thevariousplanning institutionsof the state. These fiscal and planning powersof thepresident were greatly expandedby the constitutional reformsof1968(Vi-dal Perdomo 1970). According to Vazquez Carrizosa (1986:412), 'T he es-sence ofthe 1968 reform lay inthe reduction inCongress's powerstoorganize the national budgetand to control the planning policies guidingthe economic activity of the state. W it h this reform, theprocess of over-whelming growth of presidential power reached its crowning point."Ac-cording toone ex-governmen t functionary, quoted by Vazquez Carrizosa(1986:413), The constitutional reform of 1968increased. .. the ordinarypowers of the government; its central proposition, together with that ofplanning,was tooffer the executive agile,and almost discretionary instru-ments , to deal with any imaginable situation in economic, social,and ad-ministrative policy."

The 1968 reforms initiated important changesinbudg etary powers thatremain essentially unchanged today (Arts. 210-11/1886and 349-51/1991) .The president, through the minister of finance (hacienda), preparesthe na-

15 Becauseofthe administrative natureofregulatory decrees, disputes overthepresident'suse of this power are referredtothe Council of State, the country's highest administrativecourt, rather thanto theSupreme Court.TheCouncil of State, under both constitutions,hasanappointment procedure identicaltothatofthe Supreme Court (see note4).






PRESIDENTIAL DOMINANCE I N COLOMBIA 125

tional budget and sendsitto Congress. Congress left itself onlytheabilityto reducetheamountsofvarious budg et item s;anya t tempt to increasethebudget requirestheexpress authorization ofthe presiden t. Colom bia'sCon-gress, wi thout an equivalentof the U.S.Congressional Bu dge t Office, hasby default ceded the president nearly total control over fiscal policy.Thepresident alsowasgranted in 1968 theauthority tomake contractsfor theexecutionofpub lic works projectsand toborrow monies from internationalsources. Although Congress hasaformal righ t toapprove orreject suchcontractsandloans,as the 1990s commencedithadnever madeuse ofthisfaculty.

The 1968reform also established Article 120 :14/ 188 6, which pe rm it-ted thepresident tointervenein thefunctioning ofthe C entral Bankas his"own constitutional prerogative." Exactly what this meantwasleft to judi-cial interpretation,and for several years,thepresident wasableto accumu-late more and more authority inthe management of theeconomy in thisway.Forexample,in 1972,the Supreme Court ruled unconstitutionala lawpassed by Congress that set up apub licly funded corporation toextendcreditfor agricultural development.In itsrul ing,theCo urt said thatthe lawviolated the exclusive power granted under Article 120:14/1886 for thepresident toexercise, revise,or terminate public control over personsman-aging public savings. Moreover,theCouncilofState(thesupreme adm inis-trative tribunal)in1974 ruled tha t decrees unde r Article 1 20:1 4/18 86 couldoverturn previous legislation because Congress,inenacting thearticle,hadtransferred thematterofm anagin g pub lic savings from thearenaoflegisla-tive control intothe administrative (i.e., regulatory) arena (Findley 1980).

Under thetermsofthenewcons titution , congresshasregained consid-erable control over presidential intervention in the Central Bank. Thereisno equivalent of the former Article 120:14/1886, althoughthe newconsti-tution givesthepresident exclusive authoritytoinitiate laws tha t d eterm inethe functions ofthe Central Bank (Art. 154/1991). Intervention inthema nagem ent of savingsandinvestm ent m ust take place only thro ugh frame-work laws leyes marco) that is,underpro tempore powersasdelegatedbyCongress, rather thanas an exclusive presidential prerogative.

EMERGEN CY POWERS S PRO CTIVELEGISL TIVE UTHO RITYAs noted inChapter 1, many presidents in Latin America have emergencypowers. Usually, such extraordinary authority isused only ingenuinelyextraordinary circumstances, such as insurrection, severe strikes,or naturalcalamities. These powers rarely include the ability to repeal laws or issue






126 RO NA LD P. ARCHER MAT THE W SH UG AR T

new laws. However, Colombia is a major exception, since the supposedlyextraordinary powers are used as de facto proactive legislative authority.Thereare twoforms of emergency authority that areused inColombia todecree new laws, the State of Siege (nowknown as the State of InternalCommotion)and theStateof Economic Emergency.State of Siege

Article 121/1886 gave thepresident ther ight todeclareall orpartofthe country under a state of siege, during which heholds theauthority tosuspendbut not abolish existing lawsand to emit decrees with theforceoflaw. Unlike pro tempore or regulatory decrees, these extraordinary decreeswerenotlimitedto anyparticular policy area, although they were supposedto be consistent with ' 'restoring order." Moreover, although all decreesis-sued under a state of siege automatically lapsed once the siege was liftedandanylaws suspended were automatically reinstatedatthat tim e, Congresshad little control over the substantive policies enacted during a state ofsiege.

The minimal oversight of, and the vague criteria for assuming, emer-gency powers led Colombia's presidents to make increaseduse ofthispre-rogative. From the return to civilian rule in 1958unti l the abrogationofthe former constitution inJuly 19 91 , Colombiawasunder astate ofsiegean averageof7 5 %of thet ime .Thepowers gran ted to thepresident underthe state ofsiege are such that the president, if so inclined, would appeartobe avirtual "con stitutional dic tator." Indeed,it waswith the use ofthispower that President Barco authorized a referendum on conveninga con-stituent assembly, thus bypassing the existing constitution's process forconstitutional amendment.

The constitutional terms governingthestate of siege have changed overt ime, usually limiting thepresid ent's freedom of action un dertheprovision,with the most significant changes coming with the new constitution of1991. In the original 1886 constitution, the power was truly u nlim ited,granting full powers of martial law to the president. In 1910, the firstlimitations were placed on the president's power, including clausespre-venting the president from repealing lawsandanother establishing th atthepresident and other executive branch officials would beheld responsibleforany abuses committed in theexerciseofstateofsiege powers.16It isworthnoting that the 1910amendment wasalsothe one that established the di-rect popular electionof the president, thus openingup thepossibility thatthe president would be a less reliable agent of the Congress than those16 Legislative ActNo.1of1910.Seealso Gibson (1948).






PRESIDENTIAL DOMINANCE IN COLOMBIA 127elected under the pre-1910 procedure, which employedanelectoral collegeelected on thesame basisas theCongress itself.

Before the1950s, thestate of siege was generally used as intendedthat is,mainly to comvil unres t, rather tha n as apolicy-m aking tool tobypass congressional opposition. In 1956, however, under pressure fromColomb ia's only military dicta torshipinthe twe ntieth century, the SupremeCourt said that itwas permissibleforthe p resident tousethestateofsiegepowers to attack economic aswell as political causes ofpub lic disorder.Moreover, thecourt said tha t thedecision to declare or revoke a stateofsiege wasentirely amatter of executive discretion and was notsubject tojudicial review. Theonly lim it left by that court decision was the rathervague requirement that decrees issued under the state of siege could beaimed only at restoring order andcould not override acts ofCongressforother purposes (Findley 1980).17

Given thevagueness oflimita tions placed on thestateofsiege powersby the 1956ruling and an agreement upon the return to democracy tolegalize allofficial acts tak en un de r the 195357 dictatorship, presidentscame increasingly torelyon thestate of siege powers toattack long-te rmeconomic problems.In I960 Congress amended Article 121 toprovidefora permanent congressional session during astate ofsiegeand to let eitherhouse of Congress be a plaintiff before the Supreme Court to challengespecific decrees issued pursuant to astate of siege. However, Congressre-pealed these provisionsin1968 .1 8(Alsoin19 68 , Congress created a separateand presumably more controllable stateofexception to deal with economictroubles, which we discuss later.)

Dur ing andafter theN ational Front pow er-sharing regime , stateofsiege powers were increasingly used as ameans toovercome congressionalinefficiency andimmobil ism. In a number ofcases,thestate ofsiegewascontinued long after the "crisis" used tojustify it hadreceded, andmanyof the decrees issued during these periods had little to do with the ostensiblepurpose ofdeclaring a state of siege public order. Although all thesedecrees are automatically rescinded when thestateofsiegeisdeclared over,in fact many have been incorporated into the legal code by theCongress.

17 Even this relatively favorable ruling prompted the dictator Rojas Pinilla to issueadecreereorganizing the court moretohis liking.18 One obvious reasonforthe decisioninI960tomake congressional sessions permanentduring astateofsiegeis that congress members were paidperdiem W hile congressmembers were still paid perdiem after 1968, sessions hadbecome so long thatthedistinction was probablynotso important, plusthe 1968 amendments includedasideagreement to expand congressional office staffs andtoensure each member special dis-cretionary funds {auxiliosparlamentarios).






128 R O N A L D P . A R C H E R M A T T H E W S H U G A R TIndeed, because it was a presidential prerogative to decide when to lift astate of siege, often presidents would refuse to declare an end to the siegeuntil Congress had agreed to enact into permanent law those measures thepresident deemed essential. Examples include the the Statute of Security(Extraordinary Decree 1,923 of 1978, signed by President Turbay), whichincluded a provision that placed civilians under military justice in certaincases. This provision was included within the revised Criminal Code.

As several scholars have n oted , th e pre side nt's use of state of siege pow -ers presented serious dangers for Colombian democracy, especially given thegovernment's power to restrict civil liberties and human rights severely(Umana 1985). In response to abuses of the emergency powers, Article 213/1991 places new restrictions on these powers. For example, the initial dec-laration of what is now renamed a state of internal com mo tion 19 may be forno more than 90 days, although the president may extend it for another 90days thereafter. A second 90-day extension - the last one pe rm itte d -requires the consent of the Senate. The constitution also forbids trying ci-vilians in military courts under any circumstances and states that interna-tionally recognized human rights must be upheld. The law passed in May1994 to regulate emergency powers adds two new checks: The Constitu-tional Court may of its own right overturn emergency decrees if they usurpconstitution al gu arantees; and the Congress may revoke or amen d any em er-gency decree at any time {El Tiempo, June 2, 1994). Even with these limi-tations, the state of internal commotion powers, coupled with the state ofeconomic emergency powers (discussed next), give the Colombian presi-dency formidable proactive authority, comparable to the provisional measurein Brazil, as discussed in Chapter 1 and in Mainwaring's chapter.

State of Econom ic EmergencyAs noted, presidents after 1958 came to declare states of siege to dealwith structural economic problems. Also, as noted previously, the 1968reforms to the constitution were intended to give the president discretionto intervene in the Central Bank and to make it easier for the Congress todelegate decree powers by which the president could manage the economy.However, some members of Congress, led by Senator (and future president)Alfonso Lopez Michelsen, did not believe that the new economic powerswere sufficient to prevent continued presidential reliance on Article 121/1886 for dealing with economic issues. As a result, when Congress passedthe 1968 am endm ents it created a new extraordinary power, that concerning19 Un like in the previous constitu tion, there is now a separate article that deals with a state

of external war (Art. 212).






PR ES ID EN TI AL DOM INAN CE IN COLOMBIA 129states ofeconomic emergency (A rt. 122 /188 6), which thepresident coulddeclareforperiodsnot toexceed90daysin any calendar year. During sucht ime, he could issue decrees with theforce of lawwith the intentionofrestoring "economic order." Unlike state of siege decrees, however, thesedecreesdid notlapse oncethestateofemergen cy expired. Also unl ike stateof siege decrees (before 1991), Congress could repealoramend thedecreesat any time.

The right todeclare an economic emergency, and to do sowithout aconvincing reason, hasgiven thepresident a tool that some analysts claimlets him circumven t the Congresstoim plem ent his projects (Cepeda 19 85).One example waswh en P resident Lopez Michelsen, c itin g erosion of in-comesas aresultofinflation and deficit financing, used th is powertocreatea new sales tax and to cha nge t he inco me tax lawin197 4 (Vazquez Carrizosa1986:425-27). There was little reasontobelieve that these measures woulddeal with short-term emergency conditions,asArticle 122/188 6 intended.The Supreme Court's responseto achallengetoLopez's actions was th atitsown competence to intervene in thematter wasquite limited. Thecourtmajority ruled that it lacked auth ority toreview e ither the reasons giventojustify declaring a state of economic emergency or the substance of thedecrees themselves .20Itcould review only wh ether or not all constitutionallymandated procedures hadbeen followed.21 Thecourt therefore uph eldthetax reform decrees, exceptfor onesection th at conflicted w ith aprovisionof Article 122/1886 that prohibited diminishing "the social rights of work-ers under preexisting laws" (Findley 1980).

This episode provides grist forthose who argue alongthelines of Vaz-quez Carizossa (1986:429): "Article 122we nt well beyond the intentionsof its authors and supp orters and was converted, ipso facto,into the equivalentof Article 121, giving thepresident apassport of free access to legislativepowers through theemission ofextraordinary or executive decrees." Thiscriticism appearstohave been accepted inrecent years; witness the decisionof the Supreme Court tooverturn Betancur's attem pt to use theeconomicemergency powers to institute anew taxreform in 1982. Again, in 1987,the court overturned Barco's decrees issued underastate of economic em er-gency, which, in this case, was clearly oneofa severe econom ic emergencyin theregionofUraba (Cepeda 1985).20 The court voted 19 to 5, but 4 of the dissenters ruled that thecourt shouldnothaveheard the caseatall, asitwasa"political" matter rather thanalegal question (Findley1980).21 These were, according to Article 122/1886 , the prior (nonbinding) advice of the Councilof State, the consent of all ministers, declaration ofreasons{motivos),atermination date,andasummonstothe Congress.






130 RO NA LD P. ARCHER MA TTH EW SH UG AR T

However,the logic usedby thecourtin itsvarious decisionson the useof emergency powers is consistent with an interpretation that Congress isthe ultimate arbiter of the authority held by the president, which is theargum ent that we are m akinginthis chapter.Thecourt,in its1974 decisionon Lopez'stax reform, said that becausetheCongress in 1968 left itself thepower to amend or repeal decrees issued under Article 122/1886 at anyt ime,it is up toCongressand not thecourt todecideon the "convenienceand effectiveness" ofthe decrees,aslongastheydo not conflict with specificconstitutional restrictions (Findley 1980). Thus,in the language ofprinci-palagent theory, the Congress, despite its inaction,had notabdicatedau-thorityto thepresiden t, whose actions remained under the political controlof Congress.The key question thus becomeswhy, in delegating so muchpower to the president, Congress has not summoned the will to exer-cise such political control directly. This question willbe thesubject ofSec-tion3.5.

The questionofwhy C olombian politicians would tolerate such sweep-ing presidential powers in the area of economic policy will prove to behighly relevant, since the Constituent Assembly made few changes in theprovision for economic emergency powers (Art. 215/1991). Indeed, the1991 constitution even expands the scope of these powers somewhat. Itexplicitly permits the president to decree new taxes during a state ofeco-nomic emergency, thereby indicating that measures such as Betancur'sfailed attempt to decree a tax reform would in the future be constitu-tional. However, such taxes automatically lapse when the state of emer-gency ends, which is still a maximum of 90 days from its declaration,consistent with our claim that emergency powers aregenerally not effec-tive for establishing long-term reforms. Also, the 1991 constitution ex-plicitly permits the Congress at any t ime to initiate legislation thatamends or repeals emergency economic decrees, even if these are in issueareas in which the Congress is ordinarily prevented from initiating legis-lation. It is str iking that the Constituyente continued to rely on decreepowers to address the nation's serious economic problems. However, thenew checks on this power that we have reviewed here principally thegreater legal ability of Congress to pass legislation contravening a decree- have served to move Colombiaout of thecategory of "potentiallydom-inant" presidencies and into the category of proactive presidencies (seeChapter 1). These proactive powers, we argue,are a device by whichpo-litical forces in Congress (or a constituent assembly) seek to accomplishtasks that they cannot accomplish effectively on their own. Such is theessence of a principalagent relationship.






P R E S I D E N T I A L D O M I N A N C E IN C O L O M B I A 131

PUTTING LEGISL TIVE POWERSIN CONTEXTChapters 1 and11 suggest th at presidencies w ith decree powers are mos tlikely to be found insystems in wh ich congress itself is poorly e quip pedtotackle policy matters because of internal fragmentation. Further, they notethat in some countries, congress is likely to be more conservative than pres-idents. Congress may beelected indistri cts tha t ov errepresent rural areas(which tend to be more conservative), while presidents are usually elected innational districts that perforce represent all regions equally. Unlike presi-den ts whose arsenal of legislative powers consists principally of reactive pow -ers, presidents with proactive powers can succeed in getting moreofapolicyrelative to the status quo than what congress wants. Shugart and Mainwar-ing su ggest tha t th is is so because presiden ts can use their decree power stra-tegically, such that decrees survive when they have to be turned into law bycongress. In other words, presidents cannot decree just any policy they pre-fer; congress remains a check. But presidents, using decrees, can push policyfarther from the status quo than congress would do on its own.

Two countries in Latin America clearlyfitthis pat tern: Brazil and C o-lombia. These are also the two countries where traditional clientelistic bro-kering remains strongest incongression al electio ns, especially inthe moretraditional rural areas. Furthermore, these are also the two countries thathave themost internal party com petition , asvarious broke rs com petetoelect their favored candidates. The Brazilian case isdiscussed indetailinM ainwaring's ch apter; in this chapter we have so far outlined the Colom bianpresident's legislative powers, and in the rest ofthe chapter we no tetheproblems caused by party factionalism (Section 3.4) and how this faction-alism and the disjunction between the constituent bases of the presidencyand of the Congress impose limits on presidential powers (Section 3.5).

To foreshadow the conclusions of these sections, let us recall that Col-ombian presidents before 1991 also hadaveto, imply ing that they shouldindeed have been able to getwhatever policy th ey w anted inshort,todominate Congress. However, one of our principal points inthis chapte r isthat such dominance remains more potential than real because Congressitself determines the allocation of powers under the constitution, andtheconstitution requires that most reforms, tostick over the long term , m ustbe converted into regular law. Thus, decrees are anim porta nt policyin-strument, but they are often only part of a complex negotiation with partybosses who control votes in a congress that must ultimately decide whetherto letextraordinary decrees stand as per m ane nt laws. Moreover, many re-






1 3 2 R O N A L D P . A R C H E R M A T T H E W S H U G A R Tforms require constitutional amendments, and as noted, Congress alone hasauthority over these. Only in 199091 was Congress's sovereignty over theconstitution broken; that is the subject of Section 3.6.

3 . 4 . ELECTORAL INC EN TIVE S AN DPARTY FACTIONAL ISMIf presidents in Colombia have been endowed with such formidable pow-ers to issue decree-laws to tackle economic and other problems,apuzzleemerges: Why have recent Colombian presidents been unable to carry outlarge parts of their programs, especially programs that seek to implementthe social, economic, and political changes necessary to bring the coun-try's institutions in line with aradically chan ging society? In th is sectionwe provide an explan ation based on the often-conflicting electoral incen-tives of presidents and Congress. The reason that presidents are allowed tomake substantial policy by emergency measures isthat the Congressisnot an important arena for the creation of national policy or for the rep-resentation of different policy preferences. To understand why, we arguethat one must look to the party system, which reflects clientelistic inter-ests and patronage-maintained logrolling rather than nationally orientedprogrammatic representation. Because the system is structured in suchaway as toprovide po liticians wide au tono m y from congressional p artyleadership, there is no central legislative agent with the proper incentivesand power to restrain the particularistic actions of the congressional rankand file. Therefore, Congress asawhole is often both unable and unw ill-ing to join with the executive in national policy initiatives. On the otherhand, as we discuss later, the Congress can set significant constraints onpresidential action and can restructure presidential initiatives in ways thatserve the interests of local politicians. This form of politicians' autonomyis maintained by an electoral system that encourages intense intrapartycompetition and factionalization.

Members of Congress, even if they value economic or political reform,face a collective-action problem wherein their dominant strategy is to defectfrom any reformist solution in order to ensure advancement of their ownpatronage-driven careers and those of their clients in the electorate. Thedelegation of strong but circumscribed presidential powers is a way of over-com ing the m ost debil itatin g aspects of the collective-action problem . H ow -ever, ultimately the political interests that have prevailed are those of theclientele networks that have long dominated the Colombian party system.A partial solution to this problem was found in the use of extraordinarypowers to enable the convening of the Constituent Assembly that rewrote






P R E S I D E N T I A L D O M I N A N C E INCOLOMBIA 133

the country's constitution in1 99 1. Yet even this special constitutional bodywas unwilling to change manyofthe fundam ental aspectsof the country'sparty and electoral institutions.Inthis section we (1) lookat the Colombianelectoral institutions,(2)consider how th ey provid e incentives towardpar-ticularism forpoliticians,and (3)reviewthechanges inelectoral incentivesunder therulesof the newconstitution. Then in Section 3.5 we lookatwhat effect these institutions haveon thepolicy-making powerofColom-bian presidents.

TH E ELECT OR AL SYSTEMThe electoral system used in theColom bian Congressisbasedon thecoun-try's traditionofdecentralized ConservativeandLiberal regional elites.Themost salient feature ofthe C olom bian electoral system isthat thepropor-tional-representation seat-allocation procedure isapplied ineach districttofactional lists, rather than toparty lis ts. Each list stands aloneforthe purposeof allocating seats; party is not even acriterion used in theallocationofseats. In virtually every district in every election, more than onelistrep-resenting eachof the two major parties hasbeenrun. Thereason for thisproliferation oflistsisthat party organizationsdo notcontroltheuseoftheparty label. That is,candidates, as long they meet themin imu m require-ments for registration presenting asworn statem ent of candidacyandpaying asmall feemaypresent listsof candidates under theparty nameof their choice. Thus, parties have little control over the compositionoftheir delegation in either houseofC ongress.

Moreover, theprocedureforallocating seatstoliststhe simple quotaand largest remainders rule means that the least costlyway (intermsofvotes required)to win aseat is to winwith a remainder rather thanto aimforaquota.22For example,in thedistrictofMagdalenain1990, there were258,996 valid votes cast,as can beseen in Table 3.1. Because thedistrictmagnitude was six, the simple quota was258,996/6 =43,166 . Anylistachieving this total wasguaranteed as many seatsas thenumber of timesthe quota could be divided into its vote total. Anyseatsnot filledby fullquotasareawarded , oneperlist,to thelists w ith themost remain ing votesunti lallseats have been filled. Tw o lists, bot h headed by Liberal cand idates,eachwonover 52,00 0 votes, enough foreach list to win oneseatbyquota.Neither list hadalarge enough remaindertowinasecond seat. The smallest

22 CoxandShugart (1994) have shown that because thereisno poolingor "transferring"ofvotesfrom one listtoanother,ifnolist wins enough votes to win more than one seat,the system iseffectively identicalto the single nontransferable vote,asused inJapanthrough 1993.Downloaded from Cambridge Books Online by IP 201.231.200.165 on Mon Sep 17 19:03:41 BST 2012.

http://dx.doi.org/10.1017/CBO9781139174800.004





134 RONALD P . ARCHER MATTHEW SHUGART

T a b l e 3 . 1 .A n exam ple of how the C olombian electoral system works: Results fromthe district of Magdalena in 1990

Seats byVotes Seats by quo ta Rem ainder remainder Total seatsLiberal lists1 (Abdala)2 (Menotti)3 (Ramos)4 (Aduen)Conservative lists1 (Mejica)2 (Murica)3 (Alzamora)

52,94552,55239,99133,66939,20921,63418,853

1100000

9,7799,38639,99133,66939,20921,63418,853

0011110

1111110

Note:Dist r ict ma gni tude:6;total valid votes: 258,9 96; quo ta: 43 ,16 6.winning remainder was that of aConservative list wit h 2 1,6 34 votes,orabout half a quota. As shown inTab le 3 .1 , each of four L iberal lists wonaseat. However, two of these lists "paid" over 50,000 votes per seat, almosttwo and one half times the cost in votes of the "cheapest" seat. This exampleis from just one district inone election, b ut itistypical inthat itdemon-strates how relatively small lists can take advantage of the system and winseats rather cheaply. And because parties do not control the use of their ownlabels,acand idate w ho seeks to winaseatinthis m anner cannot be deniedthe use of a party's name whether or not his or her bid issupported by theparty leadership.

P RTY F CTION LISMBecause of the cheapness of w inn ing seats and the lack of label control, therehas been a tendency for more and more of these small lists. In elections from1974 through 1990, there wasaremarkable upw ard trend inthe numberof lists that received only enough votes toelect one d epu ty orsenator. Asshown inFigure 3.1 ,inthefirst p ost -N atio nal Front election in1974,about 45% ofdeputies and 58 % ofsenators were elected from lists t ha telected no other candidate. In subsequent elections this percentage increaseddramatically, accounting foraround 80 % ofmembers in each chamberin1990. The significance ofthis trend istha t increasingly lists aresimplycampaign vehicles forasingle individu al. T hu s, even w ith nochangeinelectoral rules, intraparty competition hasincreased substan tially since






PRESIDENTIAL DOMINANCE IN COLOMBIA 13 5

100

99

Figure 5.1. Single-winner lists inColombian congressionalelections, 1974-90.1974. Insuchacompetitive environment, each headofalist must constructa network ofsupporters whodeliver votesto his or her list instead ofoneof the several others operating with the same party label inexchange forpromised services, jobs,orcash (Archer 1990). These kindsofexchangesarethe stuffofclientelismo,acommon themeofthe literatureonColombia (Diaz1986; Leal Buitrago andDavila Ladron de Guevara 1990),23 which makesvoters captive ofwhichever clientele network provides them thegoods.

In Colombia before 1991 , candidates whowere incumbents evenhadconstitutionally guaranteed funds, known as auxilios parlamentarios,by

23 The tendencyfor thetraditional partiestohave multiple listsis notanew one, eveniftherehasbeen growth inthenumberof lists electing only one member.Inthe1940sand earlier, factions ofoneorthe other party presented their own lists.TheNationalFront contributed toanincreasein party factionalization, because therequirementofparity between the partiesinevery district's representation made intraparty competitionnot justanadjunct tointerparty competition,but theonly kindofcompetition. Thus,it became less importantfor acandidatetobe tiedtoone of the traditional party leadersof hisor herregion andled to theriseofmore autonomous vote brokers (Archer 1990).The end of the National Front did not restore the traditional leaders' dominance becausethe electoral system's tendencies, reviewed earlier, maketheentryofnumerous listsofcandidates seekingtowin seats "cheaply"thenormofthe system, rather thananaber-ration (Cox and Shugart 1994; Cox and Shugart forthcoming).






136 RONALD P . ARCHER MATTH EW SH UG AR T

which they could fund pork-barrel projects that would winvotes for them(Hartlyn 1988:171-76; Leal BuitragoandDavilaLadrondeGuevara 1990).However, auxiliosparlamentarios alone were insufficient toguaranteemem-bers'electoral prospects. Camp aignsforoffice insystems suchasColombia'sweresoexpensiveand theelectoral e nvironm ent socom petitive that fundsfrom outside contributors were also needed.

Towin and hold onto dependable blocsofvoters, each can didate seeksto identify with oneofseveral national factions w ith in hisorher party.These faction leaders, who are seeking to increase support fortheir ownfactionat theexpense of othe rs, provide funds, jobs,andother pa rticularisticbenefits tovote brokers. The local candidates receive these private goodsand distr ibute them to their clientsinexchangeforvotes.AsUrrutia (1991:383) hasnoted, The political structureof the parties producesacongressand city councils mostly interested indis tr ibuting thebudget to thelargestnumber of potential clients." U nder the1991 constitution, theseat-allocation procedure remains the same; however, theSenateis nowelectedinanationwide district rather than in regional contests. Lateron, weshalldiscuss someofthe effects of this reform; however,for now thepoint isthatthe basic incentivesofmem bers remainthesameasthey always were. The reis great intraparty competition, and this makes the formation of reformistpolicy-based majorities difficult. Weelaborateonthese problem sin thenextsection.

3.5. CON GRESSIONAL INTER ESTSAN D LIMITS ON PRES IDEN TIALPOWERSAswe have noted, the constitution would appear toallow thepresidenttodominate the Congress. Moreover, the Congress, especially sincethe con-stitutional reformsof1968 ,hasfailed toexercise m anyofit sownfunctions.The Congress issimply notanationally focused ins titu tio n and hasdele-gated national policy agenda duties tothe executive, provided thatthenational agenda does not impinge upon Congress members' particularisticinterests.

In this sectionweexplainthedichotom y between formal andrealpow-ers ofthe presidency byreference to a collective-action problem thatisinherent inthe composition ofthe traditional parties. We conclude thatwhiletheColombian president hasextensive pow ers, theymay beused onlywithin arelatively cons trained space of prior consensus among the majorpoliticalandeconomic players. W he nthepresident attem ptstostep outside






P R E S I D E N T I A L D O M I N A N C E IN COLOMBIA 137

these areas of consensus, his capacity to act quickly dissipates. The growingimpasse between executive authority and defenders ofthe status q uo hascaused adispersion of power and aconseq uent erosion of presid ential au-thority. Moreover, this has occurred at a time in which increasingly complexproblems have led to growing levels of social instability and violence.

Thus , the capacity of the president tocarry o ut major changes inthepolitical, social, and economic structures of the nation is truncated. Devel-oping countries such asCo lom bia face massive socioeconom ic challengesthat place enormous pressures forchange onthestate. The inabilityoftraditional mechanisms of representation, especially parties and legislatures,to face these challenges has led to growing public expectations and demandsfor strong executive leadership. The apparently great formal powers of thepresident fuel these demands, yet executive initiatives have repeatedly stum-bled over the often informal limitations on the use of executive authority.Again, we are not arguing that the electoral system or the constitution havecausedthe problems of Colombian society, only that they have made tacklingthem moredifficultthan other arrangements probably w ould do.

The Congress maintains a blocking power over presidential attem pts toundertake major reforms: It can simply refuse topass th e necessary leg is-lation or constitution al am end m ent. W he n a reform desired by the presidentrequires structural changes that cannot beaddressed thro ugh theuseofemergency powers, the president comes up against two obstaclesintryingto win congressional passage: (1) the prob lem s of legislative coalition bu ild-ing and (2) the composition of the majority party, which leads toaseveredilemma in trying to accommodatearural-biased Congress and a pres iden tmore attuned to urban interests.

PRO BLEM S OF L EGISL TIVECO LITION BUILD INGAlthough the Colombian president has great legislative powers, when im-portant changes are proposed tothe existin g social, econom ic, or politicalstructure of the country, the president must have the support of at leastalarge part of his own party and often much of the opposition as well. Thereaso

Documents

3 - The Unrealized Potential of Presidential Dominance in Colombia