Democracy Beyond Secrecy: Assessing the Promises and Pitfalls of Collective Voting

This paper assesses collective voting as a specific mode of democratic decision-making and compares it to secret voting. Under voting, voters gather in one place decide by the show hands. We theorise two potential advantages disadvantages so defined. then draw on original survey data from largest polities practising citizen assembly Swiss canton Glarus. find that both promises pitfalls non-secret are exaggerated. Non-secret voting’s suspected – social pressure abstention do not generally materialise our sample, although for women they appear be relevant some extent. However, enabling cue-taking discursive bridging bonding equally realised limited extent only. Dieser Beitrag untersucht die kollektive Stimmabgabe bei demokratischen Entscheidfindungen im Vergleich zur geheimen Stimmabgabe. Bei kollektiven Abstimmungen versammeln sich Stimmenden an einem Ort und geben ihre Stimme offen gleichzeitig ab: Das heisst, dass jede Person sehen kann, wie anderen abstimmen, ihrerseits von allen der gesehen werden kann. Der diskutiert zwei theoretische Gefahren sozialen Druck Stimmenthaltung Chancen Möglichkeit zum Cue-Taking Stärkung des Sozialkapitals durch Bridging Bonding einer solchen Anhand Umfragedaten aus dem Landsgemeindekanton Glarus nehmen wir eine empirische Überprüfung theoretischen Erwartungen vor. Die Ergebnisse deuten darauf hin, sowohl Vor- als auch Nachteile nichtgeheimen Abstimmens überschätzt werden. vermuteten nichtgeheimer spielen unserem Sample zwar gewisse Rolle weiblichen Stimmberechtigten, lassen aber nicht generell feststellen. Umgekehrt allerdings erhofften Vorteile kollektiver nur beschränktem Mass realisiert. Cet article analyse le vote collectif comme modalité de prise décision démocratique et compare au secret. Le se caractérise par fait que les électeurs réunissent même endroit votent à la fois publiquement simultanément. Cela veut dire que, non seulement, chaque électeur peut observer comment autres, mais aussi ces derniers peuvent l’électeur exprime ses choix politiques. L’article discute d’abord deux risques pression sociale l’abstention avantages potentiels possibilité s’informer auprès autres moment du (cue-taking) renforcement liens sociaux (bridging bonding) collectif. Nous menons ensuite une statistique partir réponses tirées d’un sondage mené dans Glaris, qui pratique Landsgemeinde. Nos résultats montrent craintes espoirs vis-à-vis sont exagérés. Notre échantillon montre dangers avancés jouent un certain rôle électrices, ne être généralisés. D’un autre côté, réalisent certaines limites. Since 19th century, secrecy has been cornerstone liberal democracy (e.g. Braun 2006; Buchstein 2015; Orr Johnson 2018; Teorell al. 2017: 532; Art. 25b ICCPR). Not only does protect individuals’ political values beliefs, also ensures unhindered translation their preferences into choices. By having declare how or even whether vote, citizens can think act freely. In fact, many hold through secretly we become equal members polity Lever 2007); aggregating votes this way each every voter is treated exactly same (Schwartzberg 2014: 33). Yet pockets Western have resisted introduction secrecy, most notably assemblies US, German towns (Bryan 2004; Schaub 2016; Zimmerman 1999). Here, eligible express choice simultaneous hands physically separating opposite camps, visible all. call procedure “collective voting”, since part assembled. At time, rather than actually counting votes, oftentimes majorities simply estimated. Is compatible with basic standards free participation decision-making? prize “participating acclamatory process”, “the excitement shouting hand-waving” (Schwarzberg 32) worth cost lacking secrecy? critically discusses common view at ballot box allows decision-making. first theorising positive negative effects versus explore empirical evidence gathered informative real-world case. The goal not, however, either advocate superiority argue against (cf. Seglow 2020), nor propose alternative modalities Vandamme 2018), but merely those Gilljam 2010: 80 f.). Indeed, case study offers conditions seem conducive smooth functioning i.e. its to. bear mind when interpreting results. Collective differs three main accounts: openness (being seen others), publicness (seeing co-presence (voting time place; Table 1). assembled given chance vote. They asked decide. Thus, all present others during possible ascertain Finally, just individual expressing one’s preferences, experience. contrast, under respective wills decentrally, sequentially temporally isolated other, namely enter booth drop post ballot, if permitted. Secrecy here implies own choice, others.2 addition, result known once cast, collected counted, while centrality exclusivity now determining winners (and losers) right away. All instantly know “yes” “no”-side won, least majority clear. Studying matters several reasons. First, contributes literature benefits downsides modern societies Brennan Pettit 1990; Engelen Nys 2013; Manin Offe 1989; adding perspective hitherto largely theoretical scholarship. Second, science far matter modes dichotomy vs. 2018). Condemnation latter undemocratic anachronistic partly stems conceptual reduction. more fully below, makes difference see others, after another collectively, place. completely different allowing people officials media) access records monitor behaviour. placing level terms seeing being them perfectly line ideal 2016: 96). interest remembering though mostly small places where still use. Thousands regularly practise bindingly questions cantons 1’700 municipalities Switzerland, over 1’000 U.S., 25 Land Schleswig-Holstein (Ladner 83–85; Schmidt 2020; Distinguishing attributes clearer conceptualisation risks opportunities variants. Two extreme positions exist, emphasising negative, other implication 2007; Schwartzberg 28–39; On hand, feared create increase 1) 2) deterring who cannot afford openly participating and/or forcing differently would was protected secrecy. advocates open deny relativize 2), 3) enables cue-taking, which appreciated particularly less-informed, thus reducing epistemic gap between “hobbits” “vulcans” (Brennan 4–5). Advocates 4) visible, tangible which, encouraging inter-personal discussion enquiries about people’s observed favour integration cohesion. will address these aspects turn. (2015) concisely summarises sense me. what he calls “open voting” undue influence, immediate environment: friends, family workmates. unveiling suitable illicit transaction, adherence bargain struck beforehand monitored. Third, resources necessary exercise control strike bargains likely found rich powerful, distinctive advantage. Open cements, nothing curb, socio-political inequality. An action performed public susceptible influence agents is. Therefore, disposal better position behavior such takes observation certainly valid points key effect protection influence. powerful may always united cause. plural, developed relatively large societies, unity rare. Also, there money render actor size group ability (Olson 1965). richest enough, best organised well keep up modernisation growing electorates. might therefore no longer buying useful outcomes (Aidt Jensen 2017) open. turn, “vote buying” happen e.g. via stuffing, manipulating electoral registries (Teorell 2017; 544; Aidt or, recently, media (Orr E1a: Experiencing pronounced E1b: belonging less groups experiencing others. Being effects. (1990), behaving “discursively defensible manner”. means adopt argued lie interest, reducible commonly agreed principle(s): “People judgements put expect discursively challenged stance” 1990: 328). element stipulates I becomes extension deliberation, applying argumentative logic itself rendering decisive reasonable (Engelen 2013: 497). (2020) goes state citizens, affected my choice. Note rationale apply, need defend positions, should able required. possibility Voting pushes consider own, private interest; enhances awareness decision seriously reflected upon (Arendt 1965: 256; Barber 1984: 188; 2000: 655 695–7; Mansbridge 1983: 273; 1999: 180). Ideally, then, stimulates reflect before without however 655; 2013). Whether reality comes any close question unable answer due lack data. E2a: Overall turnout lower E2b: Compared causes larger bias privileged categories citizenry. his justification targets situations keeping records. He agnostic, consciously am defined above involves “voting eyes others” (ibid. 211), Political scientists knowledge, theorised why sub-section eclectic tilted towards merits raise analytical bar. Above all, help make 37). For (2015: exists misused, take cues “particularistic, selfish, narrow-minded” persons. Lupia McCubbins (1998), cues, senders must perceived trustworthy knowledgeable (Lupia 1998: 63). checking” (Manin 2015: 212) reasons: only, suggests, sanction use citizens’ cue decision. driven concern good much selfishness. E3a: engage cue-taking. E3b: Cue-taking used often informed groups. Another attribute distinguishes squarely co-presence, usually central town square hall. Co-presence simultaneously already saw enhance behaviour because whom ask me If happens, co-citizens subjects exerting But thinking co-citizens, objects combines roles: objects, literally, together. experience like lead increased persons me, likewise outcome, important legitimate interests beliefs consideration deciding. Bryan (2004: 286–92) argues strong tradition meetings explains fact Vermont stands out among U.S. states electing representatives socially marginalized 272–3; Parkinson 2009; Wyrsch 1927: 305). necessarily (what be) good, separate multiple good(s) co-citizens. Nevertheless, mere standing sitting together, full community, forms stark contrast approach individually institutionally, foster thoughts “for particular benefit” (Mill 2001 [1861]: 124; 39). (1990: 326), too, general creates system culture confidentiality” replaced “one respectable inquire votes.” Manin’s opportunity sharing views opinions. Similar mechanisms scholars capital Putnam 1993) place: “bridging” opposing, “bonding” concurring crucial here: mutuality grasp existing distribution variety community integral voting) part. identify like-minded opponents, start coming build trust (Buchstein 658–60). Of course, prevent telling before, But, (2013: 496) it, “makes attempt reveal similar attempts fiancé girlfriend happened Vegas. There knowing sincere” (also 658–9). Although age selfies somewhat nuanced E4a: assess experiences discussing choices positively. E4b: Discussions positively partner’s known, too. E4c: Citizens perceive discussions positively, independent partner opponent. Glarus, called Landsgemeinde decides within polity’s power require popular settles ten twenty policy year. Given federation’s strongly non-centralised nature, considerable relevance. range income property taxation education, transport environment health care law order (Linder Mueller 177). During Sunday morning year, meet debate issues tabled. happens manner, openly, hands, centrally time. elections national referendums, secretly, box. almost context expected familiar procedures. counts 40’000 inhabitants, roughly 26’000 entitled ordinary attracts estimated 7–14% electorate (Schaub Leuzinger 2018: 9). Below cantonal level, municipalities: South (ca. 7’000 voters), city 8’000) North 11’000). run semi-annual meetings, rates hover around 5%. regards, quite example developed, democracy: residents fairly wealthy educated, tertiary sector accounts biggest share economy, employment 84% institutions dominated centre-right bloc Christian-democratic, conservative parties (see Online Appendix, A1). practice That remarkable. Until eight had Landsgemeinde; abolished theirs 1848, 1928, late 1990s. six replacement (secret) ballot-box approved last cases, major reasons adduced opponents claim real alleged shortcomings were extensively adduced.3 While arguments could applied developments conflicts played eventually Landsgemeinden So undoubtedly role doing away Landsgemeinde, factor (Helg 2007: 25–42; 87–112). Furthermore, 80% ca. 2’000 municipal decisions indicates accepted parts citizenry throughout country. It precisely reason Switzerland reservation International Covenant Civil Rights (ICCPR).4 sum, favourable Its long regional local legitimacy, high living welfare lessen economic dependencies (though negligible). co-existence forces socio-economic established, attested multi-party and, historically, population centuries composed catholic protestant communities (Stolz 1968: 94–9). Last least, rule courts enough bribery pressure. absent infer works. indeed viable option contexts. investigate online conducted 800 spring 2016. When recruiting participants, canton’s administration assisted promoting using official homepage extensive email list previous marketing them. invited participate contributions mass online. Our non-randomly selected sample representative regard Appendix explicitly encouraged rarely never attend disproportionally interested politics local, referendums levels. oldest under- degree overrepresented. dispose broad varied group-specific analyses. total age, income, residency, employment. reached respondents representing entire spectrum: self-placement 11-point left-right scale rendered normal mean 5.2 corresponds parties’ shares 2014 composition nation-wide 2015 (Lutz 22). To problem deviates target population, post-stratification raking Tables A6 A10) note differences non-weighted regression estimates section. solve self-selection potentially correlated feelings Landsgemeinde.5 limits kind generalisations drawn: instance, hypothesised concerns unfounded general. reluctant interpret raw figures instead aim detect patterns linked words, deriving understanding harmful abstention, ran models determinants votes. participants ever pressured someone way, A2 precise wording). five (i.e. 2011 2015) attended. frequently held partake. reversed variables measure abstention. deal count suffer over-dispersion over-proportional presence frequent participators, binomial models.6 benefits, model People never, rarely, observe Regarding bonding, prevalence addressed depending addresser’s opinion. complement analyses results conjoint experiment built shed light motives (non)-participation Stadelmann-Steffen Gerber 2020). experiment, respondent rate hypothetical situations, consisting randomly combinations Figures 1-4 details). Participants indicate probability confronted with. propensity Note: Average Marginal Component Effect (mean 95% confidence interval). N = 870, observations 8699. 10 situations. Age35=0: Respondent 34 years old. Age35=1: 35 gender 869, 8689. Female=0: male. Female=1: female. Propensity outcome; young 235, 2350. Only below considered. “conditional losing 0” defeat expected; 1 race 2 win. 366, 3660. neighbours win company tight significant p<0.1. losi