Favourable outcomes, procedural fairness, and political support: results from a survey experiment

Fa ou able ou comes, p ocedu al ai ness, and poli ical suppo : esul s om a su ey expe imen Ped o C. Magalhães Tiago Ab il 3. 74 ci izenship in c isis in oduc ion Many s udies show ha , in hei ela ion wi h au ho i ies, people ca e no only abou ge ing hei p e e ed ou comes bu also abou he way hose ou comes a e gene a ed, especially abou he ex en o which decision-making p ocedu es a e ai . Fu he mo e, a g owing esea ch agenda has shown an in e ac i e e ec be ween ou comes and p ocedu es (B ockne and Wiesen eld 1996; Gangl 2003; G immelikhuijsen and Meije 2014; Van Dijke and Ve boon 2010; Wilking and Zhang 2017). Th ough his “p ocess-ou come in e ac ion”, he ai e he p ocedu es, he less people seem o ca e abou ou comes. Do hese aspec s o people’s ela ionship wi h au ho i ies end o gene alize o he s udy o poli ical au ho i ies and he policy ou comes hey gene a e? We belie e his ques ion is impo an o he b oade hemes o ci izenship and c isis add essed by his olume. As we ha e seen h oughou Eu ope in ecen yea s, economic c ises gene a e nega i e ou comes and hus endange he public s anding o poli ical au ho i ies. Fu he mo e, economic shocks may o ce au ho i ies o ake measu es ha , al hough po en ially necessa y and bene icial in he long un, p oduce sho - e m cos s and u he losses in suppo . Howe e , his could be less poli ically damaging i , as Mansb idge a gues, ai ness in he p ocedu es behind policy- and decision-making somehow mi iga e he de imen al e ec o deli e ing nega i e ou comes, gi ing au ho i ies “ he disc e ion o ac in ways ha , al hough no in he sho - e m in e es o ci izens, may bene i hose ci izens in he long un” (Mansb idge 1990, 175). In his esea ch no e we add ess his issue and a emp o add o he li e a u e in h ee ways. Fi s , we look expe imen ally a how bo h p ocedu al ai ness and ou come a ou abili y ma e o people’s suppo o poli ical au ho i ies, in e ms o bo h incumben app o al and (in ended) o ing. Second, we look a di e en dimensions o p ocedu al ai ness — Voice, Neu ali y, and T anspa ency — o de e mine which o hem appea s o be mos consequen ial o poli ical suppo . Finally, we es o he exis ence o a p ocess-ou come in e ac ion in he explana ion o poli ical suppo using he expe imen al me hod. Se e al aspec s emain unclea in he li e a u e es ing heo ies o p ocedu al ai ness as hey apply o he s udy o poli ical suppo . The i s conce ns he ex en o which p ocedu al ai ness and ou come a ou abili y ma e o e all o people when eac ing o poli ical and policy decisions. Based on a la ge numbe o expe imen s, Esaiasson e al. (2016) sugges ha ci izens’ a ou able ou comes, p ocedu al ai ness and poli ical suppo 75 accep ance o policy decisions is d i en much mo e by ou come a ou abili y han by p ocedu al ai ness. Ou wo k, whose expe imen al design is la gely based on hei own, asks whe he ha is also he case when we mo e om “decision accep ance” as he main dependen a iable o a i udinal and beha iou al suppo o he poli ical au ho i ies ha issues hose decisions. A second unclea aspec conce ns he dimension o p ocedu al ai ness ha migh be mos consequen ial o poli ical suppo . While many au ho s ha e s essed he impo ance o “ oice” and “neu ali y” (Rodell, Colqui and Bae 2016; Rosenbaum e al. 2017; Van C aen and Skogan 2016), a somewha neglec ed dimension has been he no ion o “ anspa ency” o “accessibili y” (bu see G immelikhuijsen and Meije 2014; Migacz, Zou and Pe ick 2017). Finally, i is also unclea whe he a p ocess-ou come in e ac ion applies when poli ical suppo is conce ned. Posi i e and nega i e esul s coexis : while Kumlin (2004) and Bøggild (2016) ind no signi ican in e ac ion e ec s be ween ou comes and p ocedu es in he explana ion o poli ical suppo , o he s udies ob ain such an in e ac ion (Magalhães 2016, 2017). me hod Ou expe imen al design closely ollows Esaiasson e al.’s (2016) model o accep ance o policy decisions, applied o a su ey expe imen . pa icipan s The su ey ook place du ing he mon hs be o e he local elec ions in Po ugal ha ook place on 1 Oc obe 2017. I was online o 58 days ( om 27 July 2017 o 22 Sep embe 2017). Online su eys allow sys ema ic access o he in o ma ion o a gi en sample, and a e an ecological me hod ha akes in o accoun : cos s ( as e , simple , and cheape ) and da a collec ion (allows access o a la ge numbe o pa icipan s in a sho pe iod o ime). Online su eys a e also less in usi e han many o he echniques (wi h ewe social desi abili y e ec s) and ha e g ea e esponse accu acy (Be hlehem and Bi ignandi 2012). We used a con enience sample; pa icipan s we e con ac ed h ough social ne wo ks (including Facebook, Linkedin, e-mail, and academic po als). The su ey was pe o med wi h he Qual ics pla o m (which p esen ed se en condi ions, andomly selec ing only one o each pa icipan , using only a single link). 76 ci izenship in c isis Pa icipan s we e also in o med abou he anonymi y o hei answe s. A alid sample o 939 esponden s was ob ained. A e age age was 42.68 (sd = 13.166), 55.1% o he indi iduals we e male. As a esul o a p elimina y sc eening, we selec ed only pa icipan s who indica ed ha hey had he igh o o e in he Po uguese municipal elec ions. ma e ials and p ocedu e Be o e he su ey wo p e- es s we e conduc ed wi h a con enience sample ( he i s wi h 40 pa icipan s, and he second wi h 41 pa icipan s), which led o he selec ion o he wo ou comes, each showing simila a e age le els o impo ance and simila dis ibu ions, bu being in e sely p opo ional. Following ha , he online su ey wi h he wo chosen ou comes and se en condi ions was conduc ed (Figu e 3.1). The su ey consis ed o an ini ial se o sociodemog aphic ques ions (gende , age, le el o educa ion, social class, esidence, esidence council, le - igh o ien a ion, and poli ical p oximi y o a pa y). Subsequen ly he su ey elici ed om esponden s hei p e e ence abou a hypo he ical in es men ou come. Pa icipan s we e asked o imagine ha he mayo o hei municipali y had o choose be ween in es ing unds in he ehabili a ion o deg aded municipal he i age o p omo ing accessibili y o municipal buildings o handicapped people. Subjec s we e hen p obed abou hei p e e ed ou come (whe he hey pe sonally suppo ed he i s o he second policy decision). Policy ou come Demog aphic da a • 2 op ions • 1 con ol & 6 condi ions • c ossed wi h policy ou comes Condi ions check Fai ness condi ions Dependen ques ions Figu e 3.1 Su ey model a ou able ou comes, p ocedu al ai ness and poli ical suppo 77 A e answe ing his ques ion, pa icipan s we e andomly assigned o one o se en condi ions (one con ol g oup and six ai ness condi ions). In he con ol condi ion, he inal decision ou come was indica ed (“p omo ing accessibili y”). Unde he emaining condi ions, we i s p imed p ocedu al conside a ions, by p esen ing subjec s wi h a lis o al e na i e decision-making a angemen s o make he policy decision. Th ee a ian s we e employed o he six emaining condi ions, exp essing he exis ence o nonexis ence o ai ness in he a ian s o Voice, Neu ali y, and T anspa ency. They in ol ed calling a mee ing o he municipal assembly whe e ci izens we e allowed o exp ess hei opinions abou he in es men (High Voice), s. making he decision immedia ely in closed execu i e mee ing (Low Voice); consul wi h independen expe s on ma e s o heal h and pa imony in he ci y (High Neu ali y), s. consul ing only wi h coun y council p esiden s o he mayo ’s own pa y (Low Neu ali y); and make all documen a ion abou he p os and cons o each in es men publicly a ailable o ci izens be o e he decision was made (High T anspa ency), s. di ulging such in o ma ion only a e he decision was made and only upon eques (Low T anspa ency). Finally, unde all o he six condi ions, as in he con ol condi ion, he ou come was e ealed (cons an o all, “p omo ing accessibili y”), bu also he p ocedu e ha was adop ed, wi h andomly assigned low o high p ocedu al ai ness in he a ian s o Voice, Neu ali y, and T anspa ency. To conclude he su ey, all pa icipan s answe ed on a 0 o 10 Like scale abou “How much do you hink he decision was ai ?”, “How much do you hink ci izens ha e had he oppo uni y o in luence he decision made?”, “To wha ex en do you hink he decision was aken in an impa ial manne ?”, “To wha ex en do you hink he decision was aken in a anspa en manne ?”, and “On he basis o he in o ma ion we ha e gi en you, how sa is ied a e you wi h he Ci y Council?”. A inal ques ion was asked “Based on he in o ma ion we ga e you, would you say ha you would eel:” wi h h ee op ions: “Mo e willing o o e o he cu en Mayo ” (2), “Mo e willing o o e agains he cu en Mayo ” (0), and “Nei he one no he o he ” (1). esul s The e a e no signi ican di e ences (x2 (6) = 3.808, p = .703) in he dis ibu ion be ween he esponse op ions in he in es men s when c ossed wi h he 78 ci izenship in c isis se en condi ions (Figu e 3.2), which ensu es he con ol o he ou come a ou abili y. Figu es 3.3 o 3.5 show a e age le els o pe cei ed Voice, Neu ali y, and T anspa ency in he di e en condi ions. Subjec s in he Low Voice condi ion show signi ican ly lowe le els o pe cei ed Voice in compa ison wi h he con ol g oup, while subjec s in he High Voice condi ion pe cei ed signi ican ly highe le els o Voice, ensu ing alidi y check o oice manipula ion. The same occu s wi h he alidi y check o Neu ali y (F (6,885) = 17.073, p < .000), and wi h he alidi y check o T anspa ency manipula ion (F (6,876) = 37.824, p < .000). Nei he he in e ac ion/mode a ion be ween he pa y o he execu i e and he choice o esponse o ou come (F (2,928) = 1.640, p = .195), no be ween he se en condi ions (F (12,913) = 1.200, p = .278) show signi ican di e ences in he a e age pe cep ion o ai ness o decision. The same occu s o he a e age le el o sa is ac ion wi h mayo ’s o ice (F (2,868) = 0.861, p = .423) and (F (12,853) = 1.021, p = .427). This sugges s ha p oximi y o he main pa y in he execu i e council o he municipali y does no o e lap o in e ac wi h expe imen al condi ions o manipula ion o jus ice o wi h ou comes. Howe e , he e a e indi idual s a is ical di e ences in he a e age pe cep ion o ai ness o decision (F (2,931) = 3.041, p = .048), a e age le el o sa is ac ion wi h mayo ’s o ice (F (2,871) = 57.039, p < .001), and he change in he in en ion o o e o he incumben mayo (x2 (4) = 54.479, Figu e 3.2 Dis ibu ion o p e e ed policy ou comes by condi ion 100%80%60%40%0% Con ol g oup Low oice High oice Low neu ali y High neu ali y Low anspa ency High anspa ency He i age ehabili a ion Accessibili y o handicapped 20% a ou able ou comes, p ocedu al ai ness and poli ical suppo 79 Figu e 3.3 Figu e 3.4 Figu e 3.5 A e age pe cep ion o oice on a scale o 1 o 10 pe condi ion (one con ol g oup and six expe imen al g oups) A e age pe cep ion o neu ali y on a scale o 1 o 10 pe condi ion (one con ol g oup and six expe imen al g oups) A e age pe cep ion o anspa ency on a scale o 1 o 10 pe condi ion (one con ol g oup and six expe imen al g oups) 101 Con ol g oup Low oice High oice Low neu ali y High neu ali y Low anspa ency High anspa ency 4.05 5.60 2.88 3.90 3.23 5.19 2.94 101 Con ol g oup Low oice High oice Low neu ali y High neu ali y Low anspa ency High anspa ency 4.83 5.72 4.34 5.56 3.83 5.74 3.94 101 Con ol g oup Low oice High oice Low neu ali y High neu ali y Low anspa ency High anspa ency 4.82 6.20 3.78 5.84 3.70 6.23 2.92 80 ci izenship in c isis oice pe cei ed ai ness ou come a ou abili y sa is ac ion wi h mayo Figu e 3.6 .42*** (.05) .18*** (.05) .58*** (.04) .10 (.06) .14** (.04) E ec s on sa is ac ion wi h Mayo ’s o ice, Voice expe imen . S anda dized coe icien s. p < .001), depending on p oximi y o he pa y o he mayo . This sugges s ha pa isanship ma e ed in people’s esponses, and ha subjec s eso ed o hei eal expe ience e en when asked abou a hypo he ical poli ical con ex . This also implies ha es ima es o he e ec s o objec i e and subjec i e ai ness condi ions on poli ical suppo need o be ob ained while con olling o pa isanship. To unde s and hose di ec and indi ec e ec s, we u ned o se e al s uc u al equa ion models (sem), based on Esaiasson e al.’s (2016) model o accep ance o policy decisions. The di e en models employed each o he se e al di e en jus ice condi ions (Voice, Neu ali y, and T anspa ency) and each o he wo dependen a iables (1) Sa is ac ion wi h mayo ’s o ice and (2) In en ion o o e. Figu e 3.6 illus a es one o he models es ima ed (Voice). Se e al ea u es a e e iden om his analysis. Fi s , whe he he decision is hypo hesized o ake place unde an “objec i e” High o Low Voice condi ion is some hing ha a ec s poli ical suppo (sa is ac ion wi h he Mayo ’s o ice), bo h di ec ly and indi ec ly. While he indi ec e ec — as media ed by he pe cei ed ai ness o decision-making — is g ea e han he di ec one, he la e is also signi ican . Second, ob aining a a ou able ou come inc eases sa is ac ion wi h he Mayo ’s o ice, bu , in he Voice expe imen , only di ec ly so. The e is no e idence in his expe imen ha ou come a ou abili y shapes pe cep ions o ai ness and, by so doing, a ec s poli ical suppo . Finally, he o al e ec o Voice is abou wice as la ge as he e ec o Ou come Fa ou abili y. a ou able ou comes, p ocedu al ai ness and poli ical suppo 81 Table 3.1 shows he esul s o all s uc u al equa ion models, by a ian o p ocedu al ai ness and by dependen a iable o poli ical suppo (sa is ac ion wi h he Mayo ’s o ice o change in in en ion o o e o he Mayo ’s pa y).1 O e all, we ind ha al hough ou come a ou abili y also a ec s pe cep ions o ai ness, he o al e ec s o objec i e p ocedu al ai ness a e always g ea e han he e ec s o ou come a ou abili y, ega dless o he objec i e ai ness condi ion manipula ed — Voice, Neu ali y, o T anspa ency — and o he speci ic ou come a iable. On he o he hand, in all analyses, mos o he e ec exe ed by objec i e condi ions is indi ec , media ed by pe cep ions o ai ness, al hough in wo condi ions — Voice and T anspa ency — objec i e ai ness also exe s a di ec e ec on poli ical suppo , ega dless o how we measu e i . An addi ional conce n o he pape is ela ed o whe he ou come a ou abili y’s di ec e ec on poli ical suppo is uncondi ional o , ins ead, i i dec eases as people pe cei ed he decision-making p ocess o be ai e . In o he wo ds, we wan o es i he media o “pe cep ion o ai ness” is also a mode a o o he ela ionship be ween ou comes and suppo . Figu es 3.7 o 3.9 show he esul s o he expe imen s o each dimension o p ocedu al ai ness looking a sa is ac ion wi h he mayo as he dependen a iable: In each o he analyses, he sign o he coe icien o he in e ac ion e m (Pe cei ed ai ness*Ou come a ou abili y) is nega i e, in he expec ed di ec ion. Howe e , in one o he expe imen s — T anspa ency — he magni ude o ha in e ac ion e m is mo e han wice as la ge as in he o he cases, and eaches s a is ical signi icance a con en ional le els. Resul s using change in he in en ion o o e o he cu en Mayo a e e y simila , whe he ea ing ha a iable as con inuous o o dinal. 1 Fo he equa ion wi h o e as he dependen a iable (0, less likely; 1, nei he no ; 2, mo e likely o o e o Mayo ), we also es ima ed a s uc u al equa ion model wi h o dinal p obi . Resul s a e e y simila .