Tackling corruption has become a big deal for many countries worldwide. Plenty of studies show that corruption and poor governance can hurt investment and economic growth (Mauro, Reference Mauro1995), weaken people’s support for politicians (Seligson, Reference Seligson2002; Anderson and Tverdova, Reference Anderson and Tverdova2003; Solaz et al., Reference Solaz, De Vries and De Geus2019), and damage trust among people (Rothstein, Reference Rothstein2011). Because of this, fighting corruption is getting more attention both in research and politics. Just in the last ten years, former leaders from places like Argentina, Brazil, Costa Rica, El Salvador, Guatemala, Israel, Italy, Nicaragua, Pakistan, Panama, Peru, Romania, Slovenia, and South Korea have been found guilty of corruption.
Even though the media and anti-corruption groups often highlight these cases, we still don’t really know how they affect public opinion. Most research has looked at the scandals themselves rather than the efforts to fight corruption. We add to this by focusing on big moves against corruption. Specifically here we look at Latin America where two ex-presidents – Carlos Menem from Argentina and Rafael Ángel Calderón Fournier from Costa Rica – were sentenced on corruption charges. Footnote 1 Using the timing of these sentences alongside national surveys from Latinobarómetro during those periods lets us use an Unexpected Event during Survey Design method (UESD) to figure out how these prosecutions affected what people think. Footnote 2
We zero in on two key things that matter for holding governments accountable: trust in institutions and political involvement. About institutional trust: past work shows that corruption hurts how much people trust both corrupt officials themselves and the institutions they belong to (e. g., Seligson 2002; Anderson & Tverdova 2003; Bowler & Karp 2004; Morris & Klesner 2010; Ares & Hernández 2017; Solé-Ollé & Sorribas-Navarro 2018). Footnote 3 But it’s less clear if fighting corruption actually rebuilds trust in bodies like prosecutors or courts (Maier 2011). Since oversight bodies are supposed to keep politicians accountable “from above,” folks might see them as standing against corrupt parts of government. Also successful prosecutions could show these agencies are capable and determined to handle abuse properly (Manion 2004). So while anti-corruption efforts might lower trust toward those involved with wrongdoing, we also want to check if they boost faith in watchdog institutions or overall belief that the state is serious about tackling corruption.
The other part we study is political engagement. Some previous research says when people get frustrated by corruption they act politically-either voting differently (De Vries & Solaz 2017) or protesting (Bauhr & Grimes 2014). But others argue participation depends on feeling like you can make a difference (Levi & Stoker 2000). If people feel nothing will change about corruption no matter what they do then they might just give up politically (Persson et al., Reference Persson, Rothstein and Teorell2013; Bauhr & Grimes 2014). With all this in mind we explore whether fighting corruption shakes off that feeling of helplessness so more citizens get involved electorally or otherwise.
1. Evidence from Argentina
On June 13th, 2013 an Argentine court sentenced former President Menem to seven years behind bars for illegal weapons sales he arranged during his first term between ‘91-‘95 with Ecuador and Croatia. This ruling was seen as a long-awaited win by prosecutors trying to hold him accountable. Footnote 4
The news grabbed lots of attention across media outlets with high public interest too. Looking at Argentina’s top newspapers Clarín and La Nación shows Menem’s trial verdict was front-page news on June 13-14. Footnote 5 Google searches tell a similar story: spikes around those days peaked sharply right when the sentence was announced.
Figure 1. Google Searches for ‘Menem’ in June 2013. The y-axis shows relative frequency of searches from Argentina looking up ‘Menem.’ Google Trends normalizes results so the highest search day scores as “100” while others show as fractions compared to it. Notice how searches jump on June 13th – when Menem’s sentence was announced.
What made Menem’s sentencing stand out is it marked one of those rare moments where justice caught up with high-level officials in Argentina after hundreds of unresolved cases involving politicians accused of graft – Clarín noted he was first prison sentence among over seven hundred such cases reported before June ’13 (Clarín June 14th ’13). Media coverage reflected this breakthrough tone almost celebratory – La Nación described it as ushering a new era focused on accountability not seen before (June 18th ’13). Footnote 6
It’s also worth noting by then Menem had lost much political influence since his presidency ended back in ‘99 so there’s little chance his sentencing date was timed or manipulated politically-no evidence suggests politics played any role shaping his case outcome either per newspaper reports we checked. That makes this case pretty clean when studying effects on public opinion around anti-corruption steps.
1.1 Analysis
Fieldwork for Latinobarómetro’s Argentine survey ran June 1-30 ’13 coinciding exactly with when Menem got sentenced June 13th allowing us define treatment vs control groups based on interview dates relative to announcement day within plus/minus seven days window. Footnote 7 Appendix S7 confirms results stay similar using bigger ±14-day window but going further isn’t possible because fieldwork period itself wasn’t longer.
Figures 2-4 present our key findings. Footnote 8 They plot average treatment effects with main text graphs while detailed tables go into Appendix S6 too showing outcomes estimated both controlling for respondent traits before event plus without controls-overall controls barely changed results.
Figure 2. Trust levels toward Courts , Parties , Congress. Estimates use controls plus city fixed effects ±7 days around Menem sentencing. Error bars show main confidence intervals at 95% ; small whiskers represent extra intervals at 90%. Ratings scaled between one-to-four points.Figure 3. Perceptions about Corruption Levels , Progress Fighting Corruption , State Capacity versus Corruption Authorities(Argentina). Effects controlled similarly within ±7-day range next to Menem sentencing date. Confidence bars follow same style ; Ratings measured via four-point scale.
< img src = "https://cambridgechronicle. ca/wp-content/uploads/2026/02/ urn: c ambr idge. o rg: i d: bina ry: 20 240604154134828 – 076…. png" class = " aop- lazy – load – image" width = " 1300 " height = " 494 " data-original-image = " /binary/version/id/ u rn: c ambr idge. o rg: i d: bina ry: 20 240604154134828- 07.. Figur e 4 Political Engagement Arg ent ina Effec ts cont rols ci ud ad fixe d effe cts wi thi n +/- days M en em s entenc ing Error ba rs repr esent confidenc e interv als Sc ale d o n sc ale fro m zer o t o one.One point three Effects O n In sti tu ti ona l Tr ust
Fir st we loo k at Tru st i n Co ng res s, Politi cal P art ies , Judiciary rated si mply o n fou r-p oint scal e wher e hig her scor es m ea n mo re posi tive vie ws.
W e foc us ed Tru st i n Co ng ress an d Pa rti es bec aus e previ ous stud ies sug ges t pe opl e ge ne raliz ing bad con duct fro m indi vidu als onto th ose inst itu tions Bowle r Ka rp Mai er W ith Me nem gu ilt conv ic tio ns pe opl e mig ht se e Co ng ress an d Par tie s negati vely Yet judg men ts abou t Judici ary cou ld be diffe ren t sinc e cour ts ha ve ro le over se eing corr upti on ef forts Man ion Posit ive pros ec uti ons ma y imp rov tr ust int hos e ins tit ut ions.
As expect ed M enem s ente nc ing lowere d tru st i n Congr ess Par tie s Fig ure two But sur prisin gly it alo so caus ed dro ps i n tru st tow ard Judi cia ry Th is run counter media ‘s happ y tone No evide nc e su gg est par tis an re act ion since Me nem wa s le ss pow erf ul pol iti ca lly b y tha t tim Part isa ns sup por ting hi m wer e smal l sectio n only six percent Ap pendix sho ws negative effect span across dem ogr aphic grou ps no re duc tion af ter removin g Ju sti cia list supp ort ers Bet ter put thi s dow nto gen era l disillu sio n wit h ju dic ial syst em Not abl ew hen ins tit utio ns un relate d lik mi lita ry ch urc h tabl es A21.
We loo ked f urt her into thre ext ra out com es Preva len ce Corr upt ion Pro gre ss Re du cin g Corr upt ion Sta te Abili ty co ded fou r poi nt scal esp ositi ve meani ng Whe ther M enem sen ten cin g ga ve opti mism sta te wo rk ers co uld fi gh t cor ru ptio nin cre ased Or de cre ase? Ou r fin ding ma tch w ith jud ici ary resu lt Ne ga tiv ity gro wn Pe op le fee l leas t pro gre ss mad e Int he sta te abi lit y sco re sim ila r tren ds app ear non sig nif ican tfiv ep ercen tle vel Altoget her corr elat ed wi th deep ere disil lus ion Ar ge nti na sta te effo rt scr ut iny.
One point thr ee E ffec ts po litical eng agement
Then wo mov eto poli tic al parti cip ati on Two ou tc omes invalid voti ng wh ether res pond ents say wil ling ness cas tin valid vote dich otom ous Fiv item dat a includ es inte nti ons joine d prot ests mot io ns poli tic ally rele vant protes ts rat ed fr om one – ten Aver age valu es trans form sca led betwee zero one inte rval.
The result s sh ow dem ob ili zat ion eve nts More pe ople pl anni ng inv ali dvot ing Exp ected ly sho wing po sitive link wi th decl ine insti tutio nal tru sts Vote incre ases b y eig ht poin ts Fig ure fou r Our findi ngs jive wit h prev ious outc ome s intu iti ons abo ut anti corru pti on ste ps raisin pes sim ism low er party conf ide nc eb oth voti ng demo ns tra tion parti cip ati on.
Additio nal tests w ere madew ide ning sam ple toa ll inte rvie wed±14dayrrange Al most no differen ce Se nsitiv ity chec ks drop prov ince indi vidualy sampl esa me consis tent heter ogeneity analy ses sho wed li ttle differe ncestab les A11-A16 App endix P ar tis anshi pseff ects remo val suppor ters PJ sma ll eff ect chan ges Non par tis aneffe ct ser ved reas suran ce Pl ace bo test Jun enahe cut off da tep lac ebla cepa dle note footnot eleve nfif teen Pla cebot est showed no effe cts earlier dates most variabl eseffectsshow ndema insame dire ctionfi gure A10(Append ix).
Two Additional Case Costa RicaOur fi rst fin din gs came wit hannoyin ques tion C annega tive impact be spe cificto Me nem ca se Just some thing unique? To test thi swelook Ca lde rôns case Costa Ri ca Sent enced Octo ber five two tho usand nine To fiv eye ars pri son fo rembe zzle ment lo an acc usa tions dur ing office ter mnineteen ninety -ninety four First trial ever fo rum pre sid ent publ icly tri ed His arrest wa sfrst fou nd years pri or Servedin jail five mon thsf oot noteonefourte enshin eles sa Import ant diffe ren ces exist betw een Calde rôna nd Me nem wh ich mak eo utrip map much vigi lant composit ein vest iga tions easier compar iso nso uc hs hypot heses.
Calder ón werent pr ote ctedby sena tor immunityunlike Me nem who cou ld avoid jail Sen cond Calde ró nwastill powerf ulwhil ei npol iti cs havin gdidn ott fade ala ongwi ththetimeli ke Me nem Datafrom Costa Ricawavelatin oba rom etroper iod Septemb er30t Octob er22weusedatecutoff Octobe rfifthkeptconsistenc yinternetsearc herspe akoftheday(Fig ure five)F ootnotefourtee nanal ysisfindsbalancegroup fair matchfor comp aredpopulati ons Table A six Footno te fif tee nand sixteen regressions av ailabl eev enwithsmall samples G enerallystatistical noise graphed but averagetrends follow Arg entina nice coherenceacrosstheo utcomes. Duet olessinfo somesubj ectspendingmissingsadlywecan only keeponsamelargesetoutcomeselectivedataavail ableforcompariso nsdiscussedb elowdes pitethisse lect ions. Obvio uslimitat ionswithexternalvalidi tymustkeepinthemindhere.
Two One Point One ResultsTrust estimates(C ourts parties congress ) Cali brated simil arly stylestatistically noisier bu talsolargerpointimpacts ont rus teyellowsignificantly reducetrust Althoughtime frame offset somewhatwiderr angeofjoboutcomesblockedmissingessomeitemsavailablilityusualcontrollingcovariates. Additionaltr usteffectsconfirmationsupplem Entarytablesandfiguresappendix. Suchgeneraldropstrustevenincl udingiudiciarymatchargentina patternrepeated. Caseagainst Calderonbigmadepublicbetterknownthanpreviousheldcasespossibleparadoxica llymakespublicfeelworseduetoknownwrongdoingratherthanhopejusticeprevails. Despitethis negativepattern, nosignificanttrustgainsevidentevenjudicialinstitutionsremainsopenquestionwhythisoccurred. Stabilityresultsappearrobustundervariousspecification checksincludingpartisanshipadjustmentsdroppinggroups, resamplingtestsplacebotestsetc. See Appendixforthoroughdiscussiondetailsfootnotes15to18.
Nextpoliticalengagementmeasures Invalidvoteagain coded dichotomously Inclinationtowardvotingprotestingconstructedfrombeliefchangeitemsrecodedbecameclearmeasuregeneralpoliticaldisengagement Noeffectinvalidvotebutstrongsignificantnegativeimpacton Inclinationtoactshowingincrease, pessimisticattitudestowardchangepossibilityresultsmirrorthosefoundargentinafairlyclose. Inshortbothcountriesseenjudicialactiondidnotboostconfidencebutmayhavedecreasedparticipationinterestpolitics. Basedonthedatasettwoexamplessuggestanti Corruptioneffortsmaybehavesimilarlimitedpositiveeffectmanydevelopingcontextsparticularlywherecorruptionsystemrootedlongtimeinstitutionallyembedded. Promisingfutureresearchcouldexplorecontextslowervulnerabilityplacespossiblymoreoptimisticeffectscriminaljusticeeffortfulfillment. Highlevelprosecutionsnowcommonplacemarkimportantfieldstudyprospectsworldwideasjudicialsystemsstrengthenfightcorruptionglobally. Thishopesparkideasnewstudiescomparebroadergeographieswithdifferentlevelscorruptionsystemsresponseimpactpoliticalcultureandsocietaltrustpatternsdueglobalnaturephenomenon.
Three DiscussionUsing infofromtwolatinamericancountrieswesoughtunderstandhowcourtdecisionsagainsttopcorruptofficialsareshapingpublictrustandparticipation. Somewhatdiscouraginglyresultsshowmoredistrustinthecourtsandinargentinamorepes simismaroundanticorruptionefforts. Thismightberelatedtolengthoftimeprocessorlackfinalitybutpositivemediareactionsmakesuchreasonunlikely. Betterexplanationperhapsthatcourtverdictsmoresotriggerrefocuspeopleattentiononthescandalthemselvesrathe..
