Welp, Yanina (2018) “Recall referendum around the world: origins, institutional designs and current debates”, in The Routledge Handbook to Referendums and Direct Democracy, Edited by Laurence Morel, Matt Qvortrup, Routledge. https://www.routledge.com/The-Routledge-Handbook-to-Referendums-and-DirectDemocracy/Morel-Qvortrup/p/book/9781138209930 Recall referendum around the world: origins, institutional designs and current debates1 Yanina Welp Center for Democracy Studies Aarau 1. Definition The recall referendum is defined as a procedure that allows the voters of a given unit to remove an elected official from office before his or her term is complete (Bowler 2004). Recall can be defined as a mechanism of direct democracy (MDD), a set of procedures allowing citizens to make political decisions directly through a vote beyond regular elections. The vote can be prescribed by a constitution or a law and, thus, be mandatory; it can be triggered without the collection of signatures by the authorities in power, be it a parliament, a government, or often a president (known as ‘top down’); or it can depend on a collection of signatures (known as ‘bottom up’) (Serdült and Welp 2012). For some authors, MDDs refer to mechanisms oriented to people’s direct intervention in policy-making, as a complement to the representative system, what excludes what excludes both recall and top-down referendums from the definition (see Papadopoulos, 1995). The first, because recall is not oriented to policies but to remove representatives, the second because top-down referendums are not initiated by the citizens,. However, such controversy does not exist in the US tradition where the referendum, initiative and recall were introduced in the same process in many states a long time ago (Cronin 1989). In our view, considering that MDDs do not refer to a political system but to a set of complementary mechanism of control, agenda setting and accountability beyond elections, recall should be easily understood as a part of the group. 1 I would like to thank Uwe Serdült for his insightful comments on this chapter. 1 Then, recall can be direct when it is activated by the people through signature collection, i.e. ‘bottom up’ initiative, and indirect when the removal of an authority by the parliament or the council has to be submitted to a vote for ratification, i.e. ‘top down’ initiative. 2. Origins The origins of the institution can be traced back to the Roman Republic, where tribunes were occasionally recalled (Qvortrup 2011). The first debates of the device in modern times are dated in the years following the American Revolution, although it was much later when the institution was regulated there, and only at the subnational level. Spivak locates its first appearance in the laws of the General Court of Massachusetts Bay Colony in 1631. Later on, ffter more than a century, there was a proposal to include recall in the Convention of 1787, but it was defeated. Federalist Alexander Hamilton lead the opposition saying that recall ‘will render the senator a slave to all the capricious humours among the people’ (quoted in Spivak 2004: 22). Not introduced into the US Federal Constitution, the recall went out of the American debate almost for a century. In Switzerland, despite of being the most well-known country for its level of decentralization and spread of MDDs, recall was only introduced in few cantons, mostly during the ‘democratic movement’ of mid to late nineteenth century (i.e. Bern in 1846, Schaffhausen in 1876, Solothurn in 1869, Ticino in 1892, Thurgau in 1869 and, later on in Uri in 1915) (See Serdült 2015). There is a debate on whether the Swiss case would have influenced the introduction of recall in the US. Apparently, the main promoter of recall in the United States, Dr J. R. Haynes, got the idea by reading ‘The City for the People’ by Parsons (1901). However, the book offers a lot of references to the Swiss example in general but does not mention recall in particular (Rappard 1912: 127). In 1892 the Socialist Labor Party and the Populist Party introduced the topic again into the US agenda, but this time from the state and city level. Finally the recall was included in the new Charter of the city of Los Angeles in 1903. The process of institutional diffusion started at the local level and spread to the currently twenty-six states having legal provisions to activate it (NCLS 2016). But the recall referendum was not restricted to liberal democracies. It was also defended from the communist side by Antonio Gramsci and Vladimir Lenin (Qvortrup 2011), clearly relating its regulation to the idea of removing ‘delegates’ more than ‘representatives’. In the same vein recall was regulated in Cuba in the Constitution of 1976, although it had been introduced earlier in the Constitution of 1951, and never used (Guzmán 2014). 2 3. Institutional designs As mentioned above, recall can be direct, when activated by signature collection, or indirect, when activated to ratify a removal already decided by the council or parliamanet. In many cases the regulation comes from the national Constitution (e.g. Poland, Japan or Peru are examples of this) while in federal countries the subnational units decide on the introduction of the mechanism (Switzerland, US, Germany or Argentina are prominent examples of this). As shown by Table 1, currently at least in 25 countries there are legal provisions to activate direct recall at national and or subnational level. Table 1: Overview of direct recall considering authorities to be removed and level of regulation Level of regulation Authorities able to be removed All elected authorities Parliament as a whole National Members of Parliament Countries (Year of introduction) Bolivia (2009), Cuba (1976), Ecuador (1998), Venezuela (1999), Taiwan RC (2003) Liechtenstein (1921) Russia, Ethiopia, Kiribati, Kirghistan, Nigeria, Liberia, Uganda, Panama and Palau Total countries 5 1 9 Executive and/or legislative Colombia (1991), Japan (1947), Poland (1991), authorities at the subnational 4 Peru (1994) and local level All elected authorities Switzerland: Uri (1915) 1 Switzerland: Bern (1846), Solothurn (1869), Parliament and/or Schaffhausen (1876), Thurgau (1869), Ticino Government as a whole (1892) Subnational Members of Parliament Canada: British Columbia (1995) 1 1 Regulated by the state, or Argentina (14 provinces) , United States of 3 province America (19 states)2, Mexico (6 states)3 Executive authorities at the Germany (four Länder)4 , Switzerland: Uri 1 local level (mayor) (1915), Ticino (2011) Source: own elaboration adapted fromSerdült and Welp 2017 and in the national Constitutions and laws. 1 In Argentina recall is regulated at the provincial level in Chaco (introduced in 1957), Chubut (1994), Córdoba (1923, 1987) , Corrientes (1960), La Rioja (1986), Rio Negro (1988), Santiago del Estero and Tierra del Fuego (1991); other provinces include it for their municipalities, namely, Entre Ríos (1933), Neuquén (1957), Misiones (1958), San Juan (1986), San Luis (1987). It is also included in Ciudad de Buenos Aires (1996) (Arques 2014). 2 In the United States is regulated in Alaska (1959), Arizona (1912), California (1911), Colorado (1912), Georgia (1975), Idaho (1933), Illinois, Kansas (1914), Louisiana (1914), Michigan (1913), Minnesota (1996), Montana (1976), Nevada (1912), Nueva Jersey (1995), North Dakota (1920), Oregón (1908), Rhode Island (1992), Washington (1912) and Wisconsin (1926) (Source: Bowler 2004) 3 In Mexico was regulated in Yucatan (1938) and Chihuahua (1997) but in both cases the National Court declared it unconstitutional in the late nineties. It is regulated in Zacatecas (1998), Oaxaca (2011), Morelos (2013), Guerrero (2014), Aguascalientes (2014) and Nuevo León (2016). See Limón 2016. 4 In Germany: Brandenburg (1993), Sachsen (1994), Schleswig-Holstein (1996), and North-Rhine Westphalia (2011) (see Vetter 2006) 3 Institutional designs diverge in relation to i) who can be removed by recall (i.e. which authorities or bodies); ii) why (i.e. for what reasons); iii) when the device can be activated (i.e. the period of activation), iv) the number of signatures; v) the time given to collect signatures; and vi) the actions taken if the authority is removed, among others. Regarding the authorities able to be removed through a recall referendum, a study developed by Serdült and Welp (2017) identified only five countries in the world in which all the elected authorities could be removed by direct recall referendum (Bolivia, Cuba, Ecuador, Venezuela and Taiwan). However, in Cuba it means that only the delegates at the lowest state level can be removed by referendum because the rest are elected by delegates and not by the people, what among other reasons, it could explain while in Cuba recall is a ‘dormant’ institution (See Guzmán 2014). The previous also shows that recall can be compatible with non-democratic systems. In Liechstenstein, under national regulation only the whole council can be removed, such as in the cantonal level in Switzerland. Then, there are cases in which the national constitution regulates recall for the subnational level. This is the case of Colombia, where only executive authorities can be removed by recall. In Japan, Peru and Poland executive and legislative authorities could be removed by a vote. In Federal countries, the regulation coming from the subnational level varies among units in the same country, as happens in Argentina (see Arques 2014), Germany (see Vetter 2006), Mexico (see Limón 2016), Switzerland (see Serdült 2015) or the USA (See Bowler 2004) (See table 1) The reasons required to activate a recall referendum express different models of representation. One is based in “dissatisfaction”, and accordingly, does not provide bases for accountability (e.g. Peru between 1994 and 2015, Ecuador between 2008 and 2010). The other is based on “programatic vote” and leads to a delegate model of representation (e.g. some American States such as Georgia, Minnesota or Washington), and sometimes empowers a public body to assess the validity of the reasons (e.g. Ecuador since 2011). The restrictions on the period in which a recall can be activated are though to give an elected authority time to prove that it is acting in accordance with citizens' expectations. California is exceptional given that the recall can be activated from the first day of an authority in office (Bowler 2004). In some countries, recall can only be activated in the second half of a term (Bolivia and Venezuela), while in others it is allowed after the first year (Peru, Ecuador and Colombia). In all the Latin American cases it is not possible to activate a recall during the final year. The number of signatures varies between cases. California displays the lowest with 12 per cent of signatures of votes casted last time for office, while in Colombia 40 per cent of signatures are requiered. Note that Peru provided until 2013 an unlimited time period to collect the necessary signatures during the mandate, while in Venezuela at the local level is fixed a surprising three day limit (i.e. the collection is organized by the electoral authority in a 4 single place with representatives from both parties the authority challenged and the challengers). A very particular model is applied in California where the recall election takes the form of an election in which the people can decide to remove the authority and in case of selecting this option also elect the replacement in the same election. In some countries if the authority is removed his or her substitute takes the position (e.g. Ecuador); in others it depend of the number of authorities removed (e.g. in Peru until 2013 only if 2/3 of the members of the local government were removed new elections were called to replace them; if less than 2/3 are removed substitutes took office). See Table 2 for some examples. Table 2: Regulation of recall in a selection of cases Country Reasons Period of activation Signatures Time to collect signatures If the authority is removed... Bolivia Programatic vote Middle of the term, before the last year 15/25/30%* 90 days New elections Colombia Programatic vote After the first year, before the last 40% 180 days New elections New elections if more than ⅓ is removed. If less, substitutes assume Peru (until 2013) After the first year, before the last 25% but a maximum of 400,000 Unlimited -- Venezuela -- Middle of the term, before the last year 20% 3 days Any time 12% voters on the last election 160 days Simultaneous election to remove or elect a new authority Malfeasance / nonfeasance a ‘serious’ crime Any time 25% voters on the last election 90 days New elections -- Any time 25% electors -- New elections California Minnesota North Dakota -- Source: own elaboration based on the regulation for each case. 5 Depending on when it takes place, new elections or substitutes take the position 4. State of the practice During the early twentieth century, the legal provisions to activate recall were confined to the subnational level (e.g. Swiss cantons, US states, Argentinian provinces) and only used with frequency in some US municipalities. Both, regulation and practices started to spread after the second world war (Japan 1947) and specially after the fall of the Berlin wall and the third wave of democracy. During the nineties the diffusion of a poorly known institution took place in different regions of the world such as Poland and Colombia (both in 1991), Peru (1993), Canada (British Columbia in 1995), Venezuela (1999) and many German Länder (See Table 1). The undisputed leadership of the US municipalities as activating recall most frequently was quickly challenged by Peru, where between 1997 and 2013 there were 5,303 activations of recall procedures against elected authorities in 45.5 per cent of all municipalities (Tuesta Soldevilla 2014, Welp 2016). Japan, which had been the second user after the USA displays an interesting pattern of usage with more than 1’200 activations (See Okamoto et al. 2014) while Poland joined the group with more than six hundred activations (Piasecki 2011) (see table 3). In Germany, despite that the institutional hurdles are relatively high in the Länder where recall can be activated, practice and success are not infrequent. However, in a similar pattern than in Peru or Poland, they take place mainly in municipalities with less than 20,000 inhabitants and, in the case of Germany, most of them are initiated by parliaments in an indirect way. Even though, few experiences in bigger cities have been observed since the nineties, such as Potsdam with 160,000 inhabitants (1998), Cottbus with 100,000 (2006) and Duisburg with almost 500,000 inhabitants (described above, toghether with the cases of Lima and Warsaw, among others) (Serdült and Welp 2017). Literature on recall relates activations of the device to three sets of quite often interrelated reasons: scandals, the impact of digital media, which add a new speed and scope for campaigning, and the leadership of political parties activating recall referendums against their opponents (Serdült 2015, Spivak 2004, Tuesta Soldevilla 2014, Welp and Serdült 2017). The study by Bowler based on 254 attempts registered in US municipalities (population over 2,500) in 1996 found that the most typical reasons to activate the device were a combination of disagreements over policy, specific charges of corruption and disagreement related to personnel changes (for example, councillors and the mayor may be subject to recall if they fire a popular police chief) (Bowler 2004: 209). In some US states, the growing number of activations has been related to the lower number of signatures required given that the percentage relates to the total number of votes cast in the previous election, which is reduced due to the increasing abstention rates (Spivak 2004). Internet spread, may also play a role facilitating fast and cheap campaigning (Spivak 2004) which coul be amplified by social digital networks diffusion. (See Anduiza et al. 2012, Breuer and Welp 2014). With more legal provisions to activate recall and more facilities to activate it 6 than other mechanisms, recall plays an important role to channel citizen’s emotions against scandals and corruption. As table 3 shows, not all the attempts end in votes, rather, most of the attempts do not end in a referendum. In the case of Germany Böhme (2008: 84) reports a period of 15 years in which only 90 of all the attempts –direct or indirect– ended with a vote, 17 of them direct. In Peru between 1997 and 2013 attempts are estimated in 20’000, while little more than a quarter finished in votes (Welp 2016). Table 3: Direct recall referendum in the world. Estimation. Country Argentina Period covered* 1923-2014 Attempts* 10 Votes 2 Bolivia 2012-2013 216 -- Canada 1995-2015 26 0 Colombia 1991-2013 169 54 Ecuador 1998-2013 786 78 Germany 1993-2008 n/d 17 Japan 1947-1999 1,250 397 Liechtenstein 1921-2015 1 -- Mexico 2012-2014 1 -- Peru 1997-2013 20,000 5,303 Poland 1990-2014 n/d 656 Switzerland 1846-2015 12 4 United States 1903-1989 ~6,000 ~4,000 Venezuela 1999-2013 167 10 Source: Own elaboration adapted from Serdült and Welp 2017. * Refers to regulation and availability of data. In Colombia 169 attempts were registered between 1991 and 2013. A total of 54 ended in votes but none of them was valid because the low level of participation (40 per cent of threshold is required to validate the election). This case exemplifies how the probability of success structures the incentives to attempt a recall, given the high thresholds to trigger a recall and to obtain the necessary quorum. The abstention rate in Colombia is the highest in Latin America (contrary to most of the countries in the region, voting is not compulsory), and it is even higher when it comes to local elections and recall referendums. In this context, the strategy of mayors facing recall is to do nothing and expect a low level of turnout. The 7 question of why the high number of signatures required (40 per cent of the electorate) often exceeds the votes later on recorded in the referendum remains open (Breuer 2011). The recall cases which have attracted most media attention are the votes against Governor of California Gray Davies in 2003; president of Venezuela Hugo Chávez in 2004; president of Bolivia Evo Morales and eight governors in 2008; and president of Romania Traian Băsescu in 2012. Less known but also interesting cases were the ones against mayor of Nagoya (Japan) Takashi Kawamura in 2011; mayor of Duisburg (Germany) Adolf Sauerland in 2011; mayor of Lima (Perú) Susana Villarán and mayor of Warsaw (Poland) Hanna GronkiewiczWaltz, both in 2013. Table 4 presents the cases. Tabla 4: Most well known cases of recall referendum (selection) Case Autorithy Date Type of recall Result California Governor Gray Davis 3 October 2003 Direct Removed Venezuela President Hugo Chávez 15 August 2004 Direct Confirmed in office Bolivia President Evo Morales and 8 Prefects (Governors) 10 August 2008 Indirect President and 7 prefects confirmed in office, one removed Romania President Traian Băsescu 29 July 2012 Indirect Removed Nagoya Parliament (promoted by Mayor Takashi Kawamura) 6 February 2011 Direct Removed Duisburg Mayor Adolf Sauerland 12 February 2012 Direct Confirmed in office Lima Mayor Susana Villarán 17 March 2013 Direct Confirmed in office 13 October 2013 Direct Confirmed in office Warsaw Mayor Hanna GronkiewiczWaltz Source: Own elaboration California, USA The most prominent case of recall in the US took place in California in 2003, when Governor Gray Davis from the Democratic Party was removed from office and replaced by Arnold Schwarzenegger (Republican Party). Davies was elected for the first time in 1999 and managed to be re-elected in 2002, with lower support than in the first election and in a race marked by low turnout and much more support to minor party candidates than usual (Kousser 2004). Quickly after the re-election the Republicans discussed the option of a recall accusing Davis of gross mismanagement of California finances by overspending taxpayer's money, and 8 threatening public safety by cutting funds to local governments (Miller 2005). The signature’s collection was organized through a volunteer effort that combined the forces of conservative radio shows and the internet and also received support from wealthy promoters (Garret 2004). The particular design of the recall referendum, combining the recall with a new election did the rest. Arnold Schwarzenegger, the former movie star turned political activist immediately became the top contender. Davis was ousted with 55.4 per cent in favour of recall. Schwarzenegger received 48.4 per cent support as a replacement among a total of seven viable candidates in a vote that split mostly along party lines (Qvrotrup 2011, Kousser 2004). For the second time in US history a governor was removed from office midterm (Serdült and Welp 2017). Venezuela Recall was introduced in Venezuela in the Constitution promoted by the government of Hugo Chávez in 1999. In August 2003, in a context of high polarization, the opposition organized first in the civil association ‘Sumate’ collected 3.2 millions of signatures. However, the National Electoral Council (CNE) rejected it on the grounds that they had been collected prematurely (before the midpoint of the presidential term, see table 2). In November 2003, once the midterm was reached the opposition collected 3.6 million signatures in a few days. The process of validation was long and highly controversial and characterized by irregularities. One of the most flagrant was the online publication of the petition signers, known as The Tascon list, to make pressure on public servants (See Kornblith 2005, McCoy 2006). The vote took place on August 15. The date is significant given that if scheduled only few days later (after August 19) a recall of President Chávez would have lead to his replacement by his vicepresident. He was confirmed in office. Bolivia The case of Bolivia is an atypical one – also named as ‘vote of confidence’ – given that the referendum was called by the President Evo Morales as a strategy to overpass the deadlock between the President (controlling the executive and the low camera) and the opposition (controlling the senate and some important territories at the subnational level, such as Santa Cruz Department) (Welp y Ruth 2017). The requisite of exceeding the percentage of voters that originally voted for the person was established, meaning that to recall Morales and vicepresident Alvaro García Linera there would have to be more than 53 per cent of the votes in favour (i.e. a majority of 51 would have not succeeded). The same rule applied for the governors of eight departments (excluding Chuquisaca, where the governor had been just elected few month before the call). The referendum was held on August 10 2008 and confirmed Morales and six prefects in office (one was removed). Romania 9 In Romania the recall referendum of president Traian Băsescu on 29 July 2012 was called after the impeachment voted by the Parliament, and accordingly was an indirect recall. A previous attempt had taken place in May 2007, when 74 per cent of voters chose to keep Băsescu in office. During his second term, again the Parliament accused the president of several charges and voted in favour of an impeachment, which had to be confirmed by the electors. Prime Minister and leader of the opposition Victor Ponta accused the President of using his power to prosecute the opposition. Because Băsescu had supported strong austerity measures he had a very low popularity, what helped recall promoters. However, the Parliament controlled by the opposition did not want to take any risks and changed the law to enable an impeachment referendum to be valid if a majority of voters voted in favour (before a minimum of 50 per cent of eligible voters voting in favour was requiered). The president was removed. Nagoya, Japan According to the literature review, a frequent reason to activate recall in Japan is founded in the difficulties to active other mechanisms of direct democracy (Okamoto et al. 2014, Takanobu 2000) producing what Takanobu (2000) describes as a shift from the primary function of the recall to a secondary one rather directed at policies. In Nagoya, a city with 1.8 mio inhabitants, the newly elected mayor Takashi Kawamura initiated a recall in order to resolve a political conflict between him and the municipal parliament. In an attempt of keeping his electoral promises, Kawamura (co-founder of the anti-tax party Genzei Nippon) wanted to cut taxes and reduce the size of parliament by half. But in the hard process of implementing it, the mayor threatened parliament with the recall as a last resort. In case of success he would step down as well and seek re-election in order to govern with a newly elected parliament more favourable to his reforms. After a highly controversial process of signature collection the referendum was organized on 6 February 2011. Parliament was recalled and new elections were held as expected by the Mayor (Serdült and Welp 2017). Duisburg, Germany The case of Duisburg (500’000 inhabitants) is connected to the Love Parade 2010 which took place in an area surrounded by rail tracks that was only accessible via tunnels. Over-crowding and the following panic among visitors trying to leave the area led to the deadly catastrophe in which 21 people died and 500 were injured in a stampede. The mayor Adolf Sauerland from the Christian Democrats (CDU) was asked to take political responsibility for having allowed the festival to take place in an unsuitable area. A first attempt to remove the mayor came in the form of a parliamentary motion to trigger the recall vote in 2010 but did not receive the required two-thirds majority (Serdült and Welp 2017). Then, through a simple majority Parliament changed the Local Government Act on 18 May 2011 in order to make this recall possible. North Rhine-Westphalia was thus the fourth Land in Germany to introduce the citizens' initiated recall procedure. Almost inmediatly after the approval of the new Act a 10 committee called 'New Start for Duisburg' started to collect signatures in order to demand a recall referendum, reached the signatures from 15 per cent of the electorate and on 12 February 2012 the mayor was recalled by 85.8 per cent of the votes (Geissel 2017). Warsaw, Poland In Poland as well as in many other countries providing for the recall as a direct democratic instrument we can observe practice moving up from the local level to medium-sized and bigger cities (Piasecki 2011: 131), and in 2013 finally up to the capital. The Warsaw recall vote of 13 October 2013 is the latest and so far most spectacular case in Poland. The process was initiated by the mayor of one of Warsaw's boroughs and rather small, local citizen groups opposing such policies as the raise of local transport tickets, the cost of rubbish collection as well as delays in the construction of the second metro line – all projects under the responsibility of Hanna Gronkiewicz-Waltz, the mayor of the political party in power also at the national level, the Civic Platform (PO) (Serdült and Welp 2017). Just after the campaign was launched the opposition party, Law and Justice (PiS), jumped on board and contributed to the collection of 134,000 signatures. Due to the required turnout threshold the recall vote was not valid (26 per cent instead of the required 29 per cent) despite the fact that a large majority of 94 percent of the participants voted against her. Later on the mayor managed to stay in power for yet another term in the 2014 local elections which she won against the candidate of PiS in the second round. (Serdült and Welp 2016) Lima, Peru As in Warsaw, the mayor of Lima Susana Villarán from the party Fuerza Social was the first female in the position. Villarán did not seem to have chances with her new party, but when the National Jury of Elections (Jurado Nacional de Elecciones, JNE) did not allow the up to then second most popular candidate to compete in the elections she became the favourite. Barely a hundred days in office, she presented the results of an audit denouncing the mismanagement of public funds by the previous mayor, Castañeda Lossio. In addition, she was determined to get a firm hold on the regulation of the local public transport system, controlled by mafias; she supported the lesbians, gays, transsexuals and bisexuals movement in a country where the conservative faction of the Catholic Church still has a strong influence; and took other controversial decisions (Vásquez Oruna 2014). In July 2012, the promoters presented the 400,396 signatures requiered to activate the device against Villarán and the whole council (40 members), with the clear intention calling for a new election if at least one third were removed. After a fierce and intense campaign the vote resulted in the recall of 22 council members (20 from Fuerza Social alone) but not of the mayor. According to the regulation, the 22 authorities removed were immediately replaced by substitutes during an interim period, until new elections took place to select the new councilors on November 24 2013. 11 5. Discussion Similar than with other mechanisms of direct democracy and probably more, recall strongly divides supporters and detractors. For Bowler (2004), the main argument in favour of recall is derived from a view of democracy that places emphasis on elected representatives as delegates and not trustees. Cronin (1989) sumarize arguments in favour and against. Advocates advance the following arguments: i) the recall provides for continuous accountability, so that voters need not wait until the next elections to get rid of an incompetent, dishonest, unresponsive or irresponsible public official; ii) it helps to check undue influence by narrow special interest; iii) it enables jurisdictions to permit their officials to serve long terms; iv) it gives an average person a reason to stay informed about civic developments between elections, v) recall offers a safetyvalve mechanism for intense feelings, and vi) it provides a sensible alternative to impeachment. Opponents affirm that i) the very premise of the recall is antagonistic to republican principles, specially to the idea of free mandate representation; ii) it makes public office less attractive to the most able individuals; iii) recall votes are divisive, disruptive, polarising, and subject to a myriad of abuses and unintended consequences; iv) recall votes are confusing, often unfair and place too much burden on the voters to keep informed between elections, and v) recall referendums are costly, unnecessary and directed against the wrong target (Cronin 1989: 133-139). Some of these arguments, whether in favour or against, are still unpproved by empirical evidence (e.g. the availability of recall and the level of public information citizen’s have or the effects of recall discouraging individuals from running for public positions); while others display different results according to different case studies (Qvortrup 2011, Welp y Serdült 2014, Vásquez Oruna 2014, Limón 2016 among others). Qvortrup (2011) has analyzed the effects of having legal provisions to activate recall on accountability focusing on the US subnational level. However, he found ‘little solid evidence that the recall changes the political landscape. It is a last resort weapon, but not one that fundamentally changes politician’s behaviour’ (Qvotrup 2011: 168). Fed by the study of the US experience, current research has discussed the role of money in activating MDDs in general and recall in particular (Garret 2004) while others have stressed on the role of low turnout and the spread of digital media on promoting an increasing number of activations (Spivak 2004). The Peruvian experience activating recall has been analyzed as a result of the erosion of political party systems and a cause of new problems of governability. According to Tuesta Soldevilla (2014) and Welp (2016), the main consequence of the extensive use of recall referendums on the local level in Peru is a state of permanent campaigning which may lead to the erosion of governability as well as democratic legitimacy. The call for new elections 12 through recall procedures has become an incentive in cases where a politician's behaviour prioritises personal interest over general welfare or institutional stability, encouraged by unstable patterns of political party competition. When local party systems are highly fragmented, parties have few political links outside the community and elections are personalized events, rather centred around a person than around political ideas or policy programmes. It is common to observe an alliance of electoral losers against the winner of the former election. It has been also common to observe that as soon as the number of activations grows the laws are ammended to prevent or make more difficult the activation. It happened in Germany, Ecuador and Peru (Serdült and Welp 2017). The case of Japan has been analysed in relation to the lack of or difficulties to activate other mechanisms of direct intervention, what promotes the use of recall as an indirect mechanism of accountability (Takanobu 2000). The inclusion of the judicial branch into the group of authorities able to be submited to recall has been controversial. For detractors, it would erode judicial independence. It was introduced in Los Angeles but no judge has been removed since 1932 (Spivak 2004). Bolivia’s 2009 constitution also introduced the judicial power when popularly elected as able to be removed through a vote. Institutional designs matters to explain the number of activations. When the reasons to activate recall are related to corruption or scandals, there is a debate on the pertinence of an election to deal with the problem. The argument is that malfeasasance or corruption should be judged by the courts. One of the most noticeable contradictions of the recall procedure emerges from the study of reasons for activation. If a public body has to evaluate the validity of the reason to withdraw an authority, then the recall does not proceed if the alleged reason is rejected, but when the reason is approved by the public body it creates a strange scenario in which what has already been judged by the authorities is submitted to a vote (Castellanos 2014, Welp 2016). As final remarks, it has been observed that the legal provisions allowing for the recall procedure have been introduced more frequently since the 1980s (see table 1) while activations in the past were restricted to small municipalities they have recently reached bigger units such as California, Warsaw or Lima or even the presidents as happened in Venezuela and Romania. Scholarly research has associated the increasing use of mechanisms of direct democracy (MDDs) in general with a crisis of representative democracy (Dalton et al. 2001). recall referendums appear as a special mechanism to channel citizens dissatisfaction because are formally oriented to increasing accountability and responsiveness of the representative system. However, the idea of ‘common people’ or individuals activating direct democracy has been challenged by the literature (Kriesi 2006, Garret 2004, Serdült and Welp 2012). In particular, the activation of a recall referendum requieres organization and resources, and in having this political parties seem to be better prepared for activating direct 13 democratic procedures and also have more incentives. Recalls are expected both in the context of scandals and political distrust and/or in the context of extreme political polarisation. Meanwhile, the erosion of traditional political party systems and the spread of social digital networks set the scene to expect a growing number of recall activations in the near future, where legal provisions exist, and more demand for including it where is not available. 6. References Anduiza, Eva, Michael Jensen, and Laia Jorba (eds.) (2012) Digital Media and Political Engagement Worldwide. A Comparative Study. Cambridge: Cambridge University Press. Arques, F. (2014) Argentina: una herramienta de los gobernados en manos de los gobernantes, in Y. Welp and U. Serdült (eds.) La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina, Serie Ciencia y Democracia, Quito: Consejo Nacional Electoral, pp. 159-186. Böhme, D. (2008) Die Abwahl von Bürgermeistern – Institution und Praxis. [Verwaltungswissenschaftliche Beiträge 38]. Bamberg: Universität Bamberg, available at http://www.econbiz.de/archiv1/2010/133361_abwahl_von_buergermeistern.pdf, (accessed 15 April 2016). Bowler, S. (2004) Recall and representation: Arnold Swarzenegger meets Edmund Burke, Representation, 40(3). pp. 200-212. Breuer. A. (2011) Obstacles to citizen participation by direct democracy in Latin America: a comparative regional analysis of legal frameworks and evidence from the Costa Rican case, Democratization, 18:1, pp. 100-134. Breuer, Anita & Welp, Yanina. Digital Technologies for Democratic Governance in Latin America. London: Routledge. Castellanos, A. S. (2014) Ecuador: la transformacion de las reglas del juego y sus consequencias (1998-2013), in Y. Welp and U. Serdült (eds.) La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina, Serie Ciencia y Democracia. Quito: Consejo Nacional Electoral, pp. 83-110. Cronin, T. E. (1989) Direct Democracy: The Politics of Initiative, Referendum and Recall, Cambridge and London: Harvard University Press. Dalton, R.; Bürklin, W. and Drummond, A. (2001) Public Opinion and Direct Democracy, Journal of Democracy, 12, pp. 141-153. Garret, E. (2004) Democracy in the wake of the California Recall, University of Pennsylvania Law Review, 153, pp. 239-284. 14 Geissel, B. (2017) Direct democracy and its (perceived) consequences: The German case, in S. P. Ruth, Y. Welp, and L. Whitehead (eds.) Let the people rule? Direct democracy in the twenty-first century, Colchester: ECPR Press. pp. 155-167. Guzmán Hernández, T. Y. (2014) Cuba: deudas pasadas y retos presentes desde la norma, in Y. Welp and U. Serdült (eds.) La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina, Serie Ciencia y Democracia, Quito: Consejo Nacional Electoral, pp. 187-205. Kornblith, Miriam (2005) The Referendum in Venezuela: Elections versus Democracy, Journal of Democracy 16, pp. 124-137. Kousser, T. (2004) The California Governor’s Recall. In Keon Chi (ed.) The Book of the States, Vol. 36. Lexington, KY: The Council of State Governments, pp. 307-315. Kriesi, H. (2006) Role of The Political Elite in Swiss Direct-Democratic Votes, Party Politics, 12(5), pp. 599-622. Limón, Walter “Revocación del mandato en México” C2D Working Paper Series 51/2016. ISSN 1662-8152 http://www.zora.uzh.ch/view/subjectsnew/C2D.html McCoy, J. (2006) The 2004 Venezuelan Recall Referendum, Taiwan Journal of Democracy, 2(1), pp. 61-80. Miller, K. P. (2005) The Davis Recall and the Courts, American Politics Research, 33(2), pp. 135-162. NCLS (2016) "Recall of State officials", 30/08/2016. Retrieved http://www.ncsl.org/research/elections-and-campaigns/recall-of-stateofficials.aspx#List_of_States_with_Process from Okamoto, M.; Ganz, N.; Serdült, U. (2014) Direct Democracy in Japan. c2d Working Papers Series 47, Aarau: Centre for Democracy Aarau (ZDA). Papadopoulos, Yannis (1995) Analysis of Functions and Dysfunctions of Direct Democracy: Top-Down and Bottom-Up Perspectives, Politics & Society 23, no. 4, pp. 421-448. Parsons, Frank (1901) The City for the People or The Municipalization of the City Government and of Local Affairs. Philadelphia, PA: C. F. Taylor. Piasecki, A. K. (2011) Twenty years of Polish direct democracy at the local level, in T. Schiller (Ed.) Local Direct Democracy in Europe, Wiesbaden: VS Verlag, pp. 126-137. Qvortrup, M. (2011) Hasta la vista: a comparative institutional analysis of the recall, Representation, 47(2), pp. 161-170. Rappard, William E. (1912) The initiative, referendum and recall in Switzerland. Annals of the American Academy of Political and Social Science 43, pp. 110–45. 15 Serdült, U. (2015) A dormant institution – history, legal norms and practice of the recall in Switzerland, Representation – Journal of Representative Democracy, 51(2), pp. 161172. Serdült, U. and Welp, Y. (2012) Direct Democracy Upside Down, Taiwan Journal of Democracy, 8(1), pp. 69-92. Serdült, U. and Welp, Y. (2017) The levelling up of a political institution. Perspectives on the recall referendum, in S. P. Ruth, Y. Welp, and L. Whitehead (eds.) Let the people rule? Direct democracy in the twenty-first century. Colchester: ECPR Press. pp. 137-154. Spivak, J. (2004) California's recall: adoption of the 'Grand Bounce' for elected officials, California History, 81 (2), pp. 20-63. Takanobu, T. (2000) 'Local Self-Governance in Japan: the realities of the direct demand system', National Institute for Research Advancement Review, 7(2): 26-29. Tuesta Soldevilla, F. (ed.) (2014) Una onda expansiva, Lima: Jurado Nacional de Elecciones. Vásquez Oruna, E. M. (2014) Cuando los vientos revocadores azotaron Lima, in Y. Welp and U. Serdült (eds.) La dosis hace el veneno: la revocatoria del mandato en Suiza, Estados Unidos y América Latina, Serie Ciencia y Democracia, Quito: Consejo Nacional Electoral: 31-56. Vetter, A. (2006) Modernizing German Local Government: Bringing the People Back, in V. Hoffmann-Martinot and H. Wollmann (eds.) State and Local Government Reforms in France and Germany: Divergence and Convergence, Wiesbaden: VS Verlag, pp. 253268. Welp, Y. and Serdült, U. (eds.) (2014) La dosis hace el veneno: La revocatoria del mandato en Suiza, Estados Unidos y América Latina. Quito: Consejo Nacional Electoral. Welp, Y. (2016) Recall referendums in Peruvian municipalities: a political weapon for bad losers or an instrument of accountability?, Democratization 23 (7), pp. 1162-1179 Welp, Y. and Ruth, S. (2017) The motivations behind the use of direct democratic mechanisms, in S. P. Ruth, Y. Welp, and L. Whitehead (eds.) Let the people rule? Direct democracy in the twenty-first century, Colchester: ECPR Press. pp. 99-119. 16