Finance

How judges think in the Brazilian Supreme Court: Estimating ideal points and identifying dimensions

Description
How judges think in the Brazilian Supreme Court: Estimating ideal points and identifying dimensions
Categories
Published
of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Please   cite   this   articlein   press   as:   Nery   Ferreira,   P.F.A.,   Mueller,   B.,   How    judges   think   in   the   Brazilian   Supreme   Court:   Estimatingideal   points   and   identifying   dimensions.   EconomiA   (2014),http://dx.doi.org/10.1016/j.econ.2014.07.004 ARTICLE IN PRESS +Model ECON381–19  Available   online   at   www.sciencedirect.com ScienceDirect  HOSTED BY  EconomiAxxx(2014)xxx–xxx Howjudgesthinkinthe   BrazilianSupremeCourt:Estimatingidealpoints   andidentifyingdimensions  PedroFernandoAlmeidaNery   Ferreira,BernardoMueller ∗ Q3  DepartamentodeEconomia–UniversidadedeBrasília,Brazil Received13December2013;receivedinrevisedform19July2014;accepted29July2014 Abstract We   use   NOMINATE   (Nominal   Three   Step   Estimation)   (Poole   and   Rosenthal,   1983,   1997)   to   estimate   ideal   points   for   all   SupremeCourt   Justices   inBrazil   from   2002   to   2012.   Based   onthese   estimated   preferences   we   identify   the   nature   of    the   two   main   dimensionsalong   which   disagreements   tend   to   occur   in   this   Court.These   estimates   correctly   predict   over   95%   of    the   votes   onconstitutionalreview   cases   in   each   of    the   compositions   of    the   Court   which   we   analyze.   The   main   contribution   of    the   paperis   to   identify   thatthe   main   dimension   along   which   preferences   align   in   the   Brazilian   Supreme   Court   is   for   and   againstthe   economic   interest   of    theExecutive.   This   issignificantly   different   thanthe   conservative-liberal   polarization   of    the   USSupreme   Court.   Our   estimates   showthat   along   this   dimension   the   composition   of    the   Court   has   been   clearly   favorable   to   the   Executive’s   economic   interests,   providingthe   setting   inwhich   the   dramatic   transformation   in   institutions   and   policies   that   the   country   has   undergone   in   last   two   decades   couldtake   place.  JELclassification: H5;H77;K10;K39 Keywords: SupremeCourt;Idealpoints;Politicalinstitutions;Lawandeconomics ©2014   National   Association   of    Postgraduate   Centers   in   Economics,   ANPEC.   Production   and   hosting   by   Elsevier   B.V.   Allrightsreserved. 1.Introduction Models   of     judicial   decision   making   forthe   American    judicial   system   have   established   aseries   of    stylized   facts   abouttheSupreme   Court   that   are   widelyacceptedeven   when   therearedisagreements   about   what   drives   those   characterizationsorhowtheyare   to   be   understood.   Itisaccepted   that   thejustices’   preferences   fall   within   asingle   conservative   to   liberaldimension.Itis   believedthat   thissingle   dimension   accounts   for   enough   of    thevariance   in    judicial   voting   that   preferencesalongthis   dimension   canbetreated   as   asufficient   metric   of     judicial   ideologyfor   the   purpose   of    theanalysis   of    a   seriesofissues   related   to   thejustices’   choices   and   behavior   (Grofman   and   Brazill,   2002;   Martin   and   Quinn,   2002;   Martin Q5  BernardoMuelleracknowledgesfundingfrom   theCNPq. Q1 ∗ Correspondingauthor.Tel.:+556181110349. Q2  E-mailaddresses: pedrofnery@gmail.com(P.F.A.NeryFerreira),bmueller@unb.br,bpmmueller@gmail.com(B.   Mueller). Q4 PeerreviewunderresponsibilityofNationalAssociationofPostgraduateCentersin   Economics,ANPEC.http://dx.doi.org/10.1016/j.econ.2014.07.004 1517-7580   ©   2014   National   Association   of    Postgraduate   Centers   in   Economics,   ANPEC.   Production   and   hosting   byElsevier   B.V.   All   rightsreserved. 12345678910111213141516171819202122232425262728  Please   cite   this   articleinpressas:   Nery   Ferreira,   P.F.A.,   Mueller,   B.,   How    judgesthink   in   the   Brazilian   Supreme   Court:   Estimatingideal   points   and   identifying   dimensions.   EconomiA   (2014),   http://dx.doi.org/10.1016/j.econ.2014.07.004 ARTICLE IN PRESS +Model ECON381–192 P.F.A.NeryFerreira,B.Mueller/EconomiAxxx(2014)xxx–xxx  et   al.,   2005;   Epstein   et   al.,2007a,b).Itis   also   believed,   though   less   consensually,   that   there   isconsiderable   drift   of    the justices’preference   over   time   (Epstein   et   al.,2007b).Althoughthe   area   of    Judicial   Politics   inwhich   thesecharacterizations   have   emerged   hasin   thepast   decadebeenthestage   of    lively   debates   and   great   advancesinthe   use   of    statistical   methods   and   formal   theory,   the   focus   of    attentionandconsequently   the   results,   have   beenlimitedalmostexclusively   to   the   USSupreme   Court.   Wetstein   et   al.   (2009)argue   that   ‘there   is   no   recent   scholarship   that   hasdemonstrated   that   a   similar   unidimensional   frameworkadequatelyexplainsthe   voting   behavior   of     judges   on   the   highcourtsof    other   countries.’   Their   analysis   of    theCanadian   SupremeCourtconcludes   that   although   an   attitudinal   model   based   on   notions   of    left–right   ideology   do   explain    judicial   behaviorinsome   areas   of    Canadian   law,   thisisnottrue   acrossall   areas,   and   they   warn   againstan   ‘American   fascination   withsimplisticnotions   of    liberal-conservative   voting   and   position   taking’   (Wetsteinet   al.,   2009:   30).Inthis   paper   we   contribute   to   the   field   of    Comparative   Judicial   Politics   by   analyzing   the   Brazilian   Supreme   Courtwiththeuse   of    formaltheoryandempirical   methods   that   have   been   developed   for   and   applied   mostly   to   the   American judicialsystem.   Political   institutions   in   general,   and   the    judicial   systemin   particular,   aresufficiently   similarto   thoseintheUS   that   those   methodsand   theory   arerelevant   for   Brazil,   yet   there   are   sufficient   differencesto   maketheexerciseinterestingandinsightful   both   for   thetesting   the   robustness   of    the   theory   and   methods   as   well   as   forgainingabetterunderstandingoftheBrazilian   institutions.   Like   the   US,   Brazilisa   presidential   systemwhere   theExecutive   faces   asimilardivision   of    power   with   a   Congress   and   a   Judiciary,   with   rulessetbyaConstitution.   Thejudicial   systemiscomposedof    several   higher   federalcourts   and   aseries   of    lower   state   courts,   withtheSupremeCourt   as   thefinalwordonconstitutional   issues.   The   Supreme   Court   iscomposed   of    11    justices,   who   are   appointed   by   the   President   and   subjecttoapproval   bya   committee   oftheSenate.   The   decisions   made   in   theSupremeCourt   frequently   have   high   profiles,   arecloselyfollowed   bythepressand   society,   and   have   greatimpacts   on   the   nation’s   destiny.Despite   allthesesimilaritieswiththe    judicial   system   in   theUS   and   many   othercountries   thereare   also   several   particularities   and   idiosyncrasiesspecifictoBrazil.   Itisthese   simultaneous   similarities   and   peculiarities   that   makeit   interestingto   apply   to   the   Braziliancontextthe   methods   of    the   Judicial   Politics   literature.Thereis   already   a   large   literature   on   the    judicialization   of    politics   in   Brazil,   which   examines   thepolitical   involvementofcourts   and   explores   how    judgesmake   decisions   and   affect   policyoutcomes   (Carvalho,   2009;   Vianna   et   al.,   2007;Castro,1997;amongothers.)   Yet,   as   Kapiszewski   and   Taylor   (2008:   752),argue   in   a   review   of    the   burgeoning   Latin Q6 American   literature   onthestudy   of    thecourt   as   a   political   actor,‘fewpieces   make   effective   useof    themany   excellentresearchtools   that   areavailable   to   carryoutqualitative   causal   analysis,   or   the   equally   useful   set   of    quantitative   analyticaltechniques.’Exceptions   for   thecase   of    Brazil   are   Leoni   and   Ramos   (2006)   and   Lanneset   al.(2012)   both   of    which   usespatialmethods   to   estimate   the   ideal   points   of    Supreme   Court    judges.   That   both   ofthese   papers   arestill   unpublishedisevidence   ofthenoveltyof    applyingthesemethods   to   the   Brazilian   Supreme   Court.Inthis   paper   weuse   a   database   of    756   Direct   Actions   of    Unconstitutionality   (ADI)    judged   by   the   Brazilian   SupremeCourt(STF)   from2002to   2012   to   analyze   the    judicial   behavior   ofthiscourt’s    justices.   ADIscan   be   initiated   byarelativelybroadlist   of    actors   (the   President,   the   Attorney   General,   Governors,   thespeakers   of    theHouse   and   theSenate,politicalparties,   theBar   Association,other   professional   organizationsand   unions)   tostrike   down   federalor   state   lawsorother   normative   acts   for   notabiding   to   theConstitution.   ThusADIs   arejudged   directly   bythe   STF   rather   thanhavingbeen   initiatedin   other   instances   ofthe    judiciary.In   an   ADItheSTF   directly   exerts   its   role   asa   constitutionalcourt.The   interest   in   thistype   ofcase   arises   because   of    the   important   implications   that   emerge   fromanADI:   thedecisioncan   overturn   alaw   produced   by   the   Executive   orby   the   Legislature,   at   both   the   federal   and   the   state   levels.Consequently,contrarytoothertypesof    cases   where   thejudicial   decision   affects   only   theparties   to   theprocess,   adecisionmade   in   an   ADI   extend   to   all   of    society. 1 A   wide   range   of    actors   canbe   involved   in   anygiven   ADI   case,suchas   the   Attorney   General   and   the   Union’s   General   Advocate   (bothof    which   mustparticipate   in   every   case)aswellas amicae   curiae ,   which   makes   for   richer   analytical   possibilities.   Also,   monocratic   decisions   or   decisions   takenbysub-sets   of    the   court   –   apossibility   in   othertypes   of    cases,   are   not   allowed   for   ADIs,   making   themmorereadilycomparablestatistically.   Finally,   an   ADIhasmany   of    thecharacteristicsof    othertypesof    casesconsidered   by   theSupremeCourt,   thus   explaining   why   they   are   almostthe   onlytype   of    action   used   in   quantitative   studiesabout   theSupremeCourt,   such   as   Leoni   and   Ramos   (2006),   Oliveira   (2008),   Jaloretto   and   Mueller   (2011),Lannes,   Desposato 1 ThisisbecausetheBrazilianjudicialsystemisstatutoryratherthancommonlaw,sothatin   regularcasespastdecisionsdonot   necessarilyconstrainfuturedecisions. 293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576  Please   cite   this   articlein   press   as:   Nery   Ferreira,   P.F.A.,   Mueller,   B.,   How    judges   think   in   the   Brazilian   Supreme   Court:   Estimatingideal   points   and   identifying   dimensions.   EconomiA   (2014),http://dx.doi.org/10.1016/j.econ.2014.07.004 ARTICLE IN PRESS +Model ECON381–19 P.F.A.NeryFerreira,B.Mueller/     EconomiAxxx(2014)xxx–xxx  3 and   Ingram   (2012)   and   Ribeiro   (2012).   This   doesnotmean   that   the   method   used   here   canonlybeapplied   to   ADIs.   In   Q7 Section   4.3,   we   use   the   same   procedure   on   a   setof    Supreme   Court   decisions   related   to   the    Mensalão   trial,   in   whichseveralpoliticians   were   accused   of    avote   buying   scheme.   This   case   fortuitously   provided   alarge   setof    comparablevotesin   which   all    justices   participated,   thus   allowingus   to   estimateideal   points   related   to   those   specific   issues.Inprinciplethe   same   could   be   doneto   other   types   ofcases   (that   is,   non-ADI   and   non-  Mensalão )   however   in   practice   itisunusualto   get   a   sufficiently   large   and   comparable   set   of    cases   that   meetthestatistical   requirements   (such   as   all    justicesvotingon   each   case).The   technique   usedin   this   paper   to   estimate   the   SupremeCourt   Justices’   ideal   points   istheNOMINATE   (nominalthree-stepestimation).   This   method   was   created   by   political   scientists   Keith   Poole   and   Howard   Rosenthal(PooleandRosenthal,   1983,   1997)   and   uses   multidimensional   scaling   to   spatially   project   preferences   using   binary   votes   as   Q8 inputs.   The   method   was   srcinally   conceived   to   study   votingbehavior   in   legislatures,   a   usefor   which   it   is   already   wellestablishedwiththe   ideal   point   scores   widely   usedby   analysts   and   even   thepress. 2,3 Although   the   data   usedin   this   paper   covers   a   fulldecadeof    Supreme   Courtdecisions,   itisnevertheless   also   true   thatitcontains   thevotes   of    only   21   specific    justices,   so   that   theresults   apply   specifically   to   theBrazilian   Supreme   Courtfrom2002   to2012.   Our   expectation   isthat   this   paper   will   help   to   disseminate   thismethod   in   Brazil   and   elsewhere   andinducemore   studiescovering   otherperiods   and   domains,   thus   contributing   to   assess   theexternal   validity   of    ourresults.Thefirst   major   output   fromthe   paper   istheidealpointestimates   for   all   Supreme   Court   Justices   from2002   to   2012.BecauseBrazilian    justices   facemandatoryretirement   atage70thereisaquicker   turnover   of    theCourt’s   compositionthanintheUS.   Consequently   we   estimate   ideal   points   for   nine   different   compositions   ofthe   Court   over   thisperiod.Thelist   of    periods   and   thecomposition   of    the   court   in   each   periodareshownin   Table   1.   For   each   period   we   provide   anidealpoint   for   each    justicealong   two   dimensions,   which   canbe   conveniently   presented   in   a   unitary   circle.   Weare   alsoableto   estimate   ideal   points   forthe   Attorney   General   and   forthe   Union’s   General   Advocate,   asthey   must   givetheirpositionon   every   ADI.The   former   acts   as   acustodian   ofthe   law,   while   the   latter   is   in   charge   of    representing   theExecutiveintheSupreme   Court.   This   characteristic   of    the   Brazilian   Supreme   Court   allows   usto   morerigorously   makeinferencesof    the   STF’s   position   relative   to   theExecutive   and   theLegislature   (or   at   leasttheLaw)   and   may   eventuallybeuseful   for   analysis   of    strategic   behavior.   As   ameansto   testthegoodness   of    fit   of    the   estimated   idealpoints   theyareusedto‘predict’   theactual   votes   in   eachperiod.The   accuracy   varies   from   one   period   to   thenext,   but   more   than   90%ofthevotes   are   always   correctly   ‘predicted’,   indicating   that   theidealpointwe   estimate   are   a   goodmeasure   to   locatethejustices’   policy   preferences.Perhaps   the   major   analytical   contribution   of    this   paper   istheidentification   ofthe   dimensions   along   which   preferencesaligninthe   Brazilian   Supreme   Court. 4 For   the   USitiswell   established   that   preferences   in   the   Supreme   Court,   asinCongress,   canbe   wellportrayed   as   aligning   along   a   single   conservative   to   liberaldimension.   This   reflectsthepolarizationof    American   politics   that   ispresent   in   all   its   political   institutions,   fromelectoralrulesto   political   partiestothe   working   of    Congress,   pervading   even   the    judiciary.   Given   that   Brazil   has   amoreproportionalthan   majoritariansystema   similar   polarization   exists   but   is   more   nuanced   and   fragmented.   These   characteristics   are   reflected   in   ourresults,aswe   found   that   including   aseconddimensionprovided   useful   information.   Contraryto   the   US,   neither   of thedimensions   can   be   depictedas   capturing   a   leftto   right   or   aliberalto   conservative   alignment   of    preferences.   Ouranalysisindicates   insteadthat   the   first   and   most   important   dimension   captures   disagreements   between   the   Justices   overlawsthat   are   inthe   Executive’s   direct   interest,   especially   when   relatedto   economic   issues.Asthedata   covers   aperiodinwhich   Brazil   underwentahighly   transformative   process   of    institutional   strengthening   and   economic   reform(Alston   Q10 et   al.,   2013),   inwhich   theCourt   played   an   activerole   as   apoliticalactor,   itisan   important   finding   that   thisdimension 2 ForaprimerontheNOMINATEmethodseeEversonetal.   (2009).For   aprimerontheuseofa   softwarepackageforcalculat- Q9 ingNOMINATEscoresin   R,seePooleetal.   (2011).Forarecentmanualsee   Lo(2012)‘UsingW-NOMINATEin   R’availableathttp://cran.r-project.org/web/packages/wnominate/vignettes/wnominate.pdf . 3 BecausetheSupremeCourtisamuchsmallergroupthanLegislatures,forwhichthistechniquehasbeenmorefrequentlyapplied,theremaybestatisticalissuesknownas‘micro-committeeproblems’inextendingtheanalysisto   courts.Newtechniqueswhichmaybebettersuitedtocourtshavebeendevised,suchastheMartinandQuinnscoresemployingMarkovchainMonteCarlo(MCMC)methodstofitBayesianmeasurementmodelsofjudicialpreferences(MartinandQuinn,2002;Bailey,2007).WeintendtocomparetheNOMINATEresultstothoseofothermethodsforBraziliandatainfuturework. 4 Thestatisticalprogramyieldstheideal   pointsbutdoesnotprovideanyinformationonwhatisthecontentofeachdimension.Identifyingwhatoverarchingissuesorcharacteristicsexplaineachdimensionmustbe   inferredbytheresearchbycomparingtheestimatedidealpointsto   specificsalientvotesand/orvoters.Identifyingdimensionsisthus,inpart,anart. 7778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117  Please   cite   this   articleinpressas:   Nery   Ferreira,   P.F.A.,   Mueller,   B.,   How    judgesthink   in   the   Brazilian   Supreme   Court:   Estimatingideal   points   and   identifying   dimensions.   EconomiA   (2014),   http://dx.doi.org/10.1016/j.econ.2014.07.004 ARTICLE IN PRESS +Model ECON381–194 P.F.A.NeryFerreira,B.Mueller/EconomiAxxx(2014)xxx–xxx  Table1SupremeCourtJusticesfrom2002to2012.SupremeCourtJudgeNominatingPresidentDateMoreiraAlves Geisel20/06/1975SydneySanches Figueiredo31/08/1994SepúlvedaPertence Sarney17/05/1989CelsodeMelloSarney17/08/1989CarlosVellosoCollor13/06/1990MarcoAurélioCollor13/06/1990IlmarGalvãoCollor26/06/1991MaurícioCorrêaItamar   15/12/1994NelsonJobimFHC   15/04/1997EllenGracieFHC   14/12/2000GilmarMendesFHC   20/06/2002CezarPelusoLula25/06/2003AyresBrittoLula25/06/2003JoaquimBarbosaLula25/06/2003ErosGrauLula30/06/2004RicardoLewandowskiLula16/03/2006CarmenLúciaLula21/06/2006MenezesDireitoLula05/09/2007DiasToffoliLula23/10/2009LuizFuxDilma03/03/2011RosaWeberDilma19/12/2011 Source :   SupremoTribulaFederalhttp://www.stf.jus.br/portal/composicaoPlenaria/composicaoPlenariaAnterior.asp. has   been   the   major   guiding   principle.   Similarly,   given   that   the   estimated   idealpoints   on   thisdimension   capture   theJustices’preferences   on   federal   economic   policy   they   canbe   used   by   otherresearchers   to   conductanalyses   ortesthypotheseswhere   the   Supreme   Court   Justices’   preferences   arerelevant.Onthe   second   dimension   we   identified   preferences   relating   to   therole   of    theStates   ofthe   federation   and   theirautonomyto   create   lawsand   establish   policies.   Scree   plots   of    eachperiod   indicated   that   thoughthesecond   dimensionexplainssignificantly   less   of    thevariance   than   does   thefirst   dimension,   itnevertheless   has   enough   explanatory   powertocompensate   for   theloss   of    parsimony   entailed   in   using   two   rather   thanone   dimension.   The   thirddimension,   on   theotherhand,   provides   insufficient   marginal   explanatory   power   to   warrant   its   consideration.Afterestimating   theideal   points   and   identifying   thedimensions   we   provide   four   differentapplications   using   thesemeasuresto   analyze   theworking   and   different   characteristicsof    theBrazilian   Supreme   Court.   These   applications   servebothto   show   the   range   of    possibilities   of    these   methods   for   studying   Judicial   Politics   in   Brazil   and   other   countriesbesidestheUS,   as   well   as   providing   insights   related   to   thefunctioning   of    political   institutions   in   Brazil.   The   firstapplicationisto   identify   the   Court’s   median   Justice   in   eachof    theperiods.   Recognizing   which   Justice   is   the   medianvoteriskey   for   understanding   thecourt’sdecisions,   asunder   some   fairlygeneral   conditions   the   median   Justice’s   policychoicewill   alsobe   thechoiceof    theCourt   majority.   Yet,as   Martinet   al.   (2005:   1279)   argue,itisnotalways   obvioushowto   locate   and   identify   the   Court’s   median.   As   we   consider   two   dimensions   it   ispossible   forthemedian   to   be   adifferentjustice   in   each   dimension.   In   mostperiods   however   the   median   in   one   dimension   coincided   with   that   in   theother.Another   interestingresult   wefind   is   that   the   median   Justice   in   mostperiods   istheJustice   that   hasmostrecently joinedthe   Court.   This   ‘rookie   effect’suggests   that   the   President   has   been   able   to   closely   influence   theCourt’s   decisionsinour   sample   period.The   secondapplication   also   centerson   the   notion   of    a   median   Justice.   We   use   the   estimatedidealpoints   to   consider   the   possibilities   for   thecurrentPresident   (Rousseff)   toremake   theCourt   in   the   next   years,consideringthe   factthat   several   Justices   will   be   forcefully   retired(70yearsof    age)   in   thenext   2years.   This   followstheexercise   performed   in   Martin   et   al.   (2005)   that   used   estimated   Supreme   CourtJustices’   ideal   points   to   assess   theconventionalviewthat   existed   at   thetime   (2005),   that   President   George   W.   Bushwould   be   ableto   remake   the   Court.Thethird   application   isto   consider   whatthe   estimated   ideal   points   imply   about   thevalidity   of    three   different   theoriesofjudicial   behavior;   attitudinal,   legaland   strategic.   The   main   conclusion   is   that   thoughdifferent   aspectsof    each   of  118119120121122123124125126127128129130131132133134135136137138139140141142143  Please   cite   this   articlein   press   as:   Nery   Ferreira,   P.F.A.,   Mueller,   B.,   How    judges   think   in   the   Brazilian   Supreme   Court:   Estimatingideal   points   and   identifying   dimensions.   EconomiA   (2014),http://dx.doi.org/10.1016/j.econ.2014.07.004 ARTICLE IN PRESS +Model ECON381–19 P.F.A.NeryFerreira,B.Mueller/     EconomiAxxx(2014)xxx–xxx  5 thesetheories   are   confirmed,   other   are   not,   and   in   theend   we   are   unableto   refute   any   of    the   explanations   in   favor   of another.Thefinal   application   isalookat   some   important   individual   casesusing   the   estimated   ideal   points.   The   NOMINATEmethodnot   only   locates   each   Justice,   theAttorneyGeneral   and   theExecutive’s   Advocate   in   a   unitarycircle   composedoftwo   dimensions,   but   italso   situates   for   each   individual   case   a   dividing   line   that   represents   the   estimation   of    howtheseactors   broke   down   in   favor   or   againstin   that   case.   These   diagrams   makevisuallyexplicitthedegree   to   whichtheideal   points   explain   thevotes   in   each   caseand   also   showwhere   it   gets   the   votes   wrong.One   purpose   of    thisapplicationis   to   showhow   this   method   canbe   usedby   researchers   to   gaininsights   in   specific   seminal   cases.   One   setof thecases   we   analyzeiscomposed   ofthe   votes   in   therecent   high-profile    Mensalão   scandal,   where   highlevel   officialsandpoliticians   were   triedand   some   condemned   for   a   vote-buying   scheme   that   was   bustedin   2005.   The   importanceof this   case   lies   in   thefactthat   the   Brazilian    judiciaryhashistorically   beenloathto   punish   therich   and   powerful,optinginsteadfor   an   excessive   concern   in   guaranteeing   fundamental   rights   and   due   process,   which   the   accused   haveoftenbeen   able   tomanipulate   intheir   favor,leading   tofrequent   impunity.   The    Mensalão   trial   broke   all   expectations   byharshlycondemning   powerfuland   well-connected   politicians   and   willprobably   become   a   watershed   in   theSupremeCourt’shistory.   Itthus   warrants   aclose   examination   of    the   votes   using   ourideal   pointestimates. 2.   Estimation   of    BrazilianSupreme   CourtJustices’   ideal   points 2.1.    ADIs    –    directactionsof    unconstitutionality TheDirect   Actions   of    Unconstitutionality   (ADIs)arethe   main   instrument   of     judicial   reviewin   theBrazilian    judicialsystem.Inthese   actions,   theBrazilian   Supreme   Court   decides   over   theconstitutionality   of    lawsor   norms,withrespecttothe   Constitution.   Anaction   can   beapprovedor   rejected:   ifapproved,thenormisdeclared   unconstitutional.   AnADIcannot   bedecided   individually   byonlyone   Justice,   but   by   allofthemtogether.A   normwill   be   consideredunconstitutional   when   at   least   sixof    the   eleven   Justices   vote   for   it.Abroad   range   of    actors   canfile   an   ADI.The   Court   cannot   act   ex-officio ,   so   the   claim   of    unconstitutionalitymustbe   brought   byone   of    these   actors.   The   list   includes   thePresident,   the   Attorney   General,   StateGovernors,theExecutiveCommittees   of    the   Houseor   the   Senate,   political   parties,   theBar   Association,   union   confederations   andnationalassociations   or   classentities.   Federal   and   state   norms   or   laws   canbe   reviewed,   from   any   of    thethree   branches:Executive,Legislative   and   Judicial.   Therefore,   a   wide   range   ofinterests   canbe   involved   in   an   ADI.Accordingto   theConstitution,   everyADI   musthave   theparticipation   of    theAttorney   General   and   theAdvocateGeneralof    theUnion   (the   Brazilian   equivalent   of    the   Solicitor   General   of    the   United   States).   The   Attorney   General(knownas‘PGR’)   officiates   as   custos   legis   and   issuesan   opinion,   agreeing   or   disagreeing   with   theaction.   The   Courtmayor   may   not   followthis   opinion.   Onthe   otherhand,   theAdvocate   General   of    the   Union   (known   as   ‘AGU’)   should always ‘defend’   the   law   in   question,   arguing   forthe   rejection   of    theaction,   even   if    this   law   is   astate   law.   However,thiscompulsory   role   of    the   Advocate   General   hasactually   been   abandoned   in   recent   yearsand   the   Advocate   Generalfrequentlyarguesfavorably   to   the   approval   oftheaction   and   the   unconstitutionality   of    the   law.   Weinterpret   thatbecauseof    thisdiscretionary   behavior   of    the   Advocate   General   of    theUnion,   herepresents   the   interestsof    thefederalgovernmentin   theADIs.   Both   theAttorney   General   (PGR)and   theAdvocate   General   of    the   Union   (AGU)   areappointedbythePresident.   The   difference   isthat   while   theAdvocate   General   isan   unconstrained   choice   of    thePresident,   whocannominate   or   replace   the   Advocate   at   will,   theAttorney   General   must   beconfirmed   by   the   Senate   and   has   a   fixedterm.Since   the   Attorney   General   and   the   Advocate   General   of    theUnion   participate   of    every   case,   idealpoints   of    theseactorswere   also   estimated,   creating   useful   points   of    reference   to   understand   theJustices’   estimated   ideal   points. 2.2.Data We   use   a   database   of    756   Direct   Actions   of    Unconstitutionality   (ADI)    judged   bythe   Brazilian   SupremeCourt   (STF)fromJune   2002   to   March   2012.   This   sampleincludes   only   finaldecisions   (injunctionsdecisions   werenotconsidered)anddecisions   where   theunconstitutional   claim   wasactually   analyzed   (decisions   regardingonlyprocedure   lawwerenotconsidered).Data   wasmainly   compiled   frominformation   about   theactions   available   online   on   theCourt’s   webpage,andalso   from   the   STF’sarchive. 144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x