Abstract
The 2014 Sunflower Movement led to rising political participation among young Taiwanese. Hence, opposition parties and civic groups created programs to support young candidates running in the village chief elections. Compared with the 2010 election, however, fewer young challengers ran in 2014, and they received fewer votes and won fewer seats. Propensity score matching shows that the presence of young candidates on ballots did not increase turnout. However, young candidates affected the election indirectly: young, new candidates attracted more votes from incumbents than from challengers and therefore decreased the incumbent re-election rate.
Keywords
Introduction
The March 2014 Sunflower Movement was undoubtedly one of the most important events in Taiwan’s recent political development. Hundreds of thousands of protesters occupied Parliament (the Legislative Yuan) and the surrounding blocks, requiring authorities to reconsider the Cross-Strait Service Trade Agreement (CSSTA) and to decelerate cross-strait economic cooperation. This peaceful occupation successfully stopped the ruling party’s plan to pass the CSSTA and won support for more supervision of the agreement (Ho, 2015).
A salient feature of the Sunflower Movement is the age of its participants – they are mostly members of the young generation. According to an on-site survey conducted during the movement by Chen (2014), the average age of the participants was 28, and 67% of participants were younger than 30. Moreover, 56% of the participants were students, although nearly half of the protesters had graduated. Another online survey conducted on the third day of the movement (PollcracyLab, 2014) showed that 81% of subjects under 30 supported the occupation, compared to 32% of the subjects over 50. Therefore, the Sunflower Movement is characterized as a movement of young people rather than of students exclusively.
In a democratic system, the legitimacy of power and policies is provided by general elections. Since the Sunflower Movement was vividly imaged as the young generation, young people were also expected to play an essential role in the local election later that year: they had come out to protest, so they might be willing to run in local elections against members of the ruling party, Kuomintang (KMT). Indeed, researchers have noted that the grievances of the young generation actively contributed to KMT’s losses in the 2014 local elections (Hsieh, 2015; Subba, 2016).
Beyond getting out the vote, young Taiwanese could influence the upcoming local elections directly by running for office. Notably, the major opposition party, the Democratic Progress Party (DPP); the small, pro-independence party, the Taiwan Solidarity Union (TSU); and numerous civic groups actively encouraged young people to run for office in the 2014 village elections. 1 These groups believed that young candidates could overturn local politics by means of the framing effect of the Sunflower Movement.
Given the atmosphere and resources favoring the young generation, its role in the upcoming elections was an open question. Would the sunflower also bloom in local politics? Most previous studies of the effect of the Sunflower Movement on Taiwan politics have focused on the movement itself, rather than on its effects on the next local elections, which occurred seven months after the end of the movement (Ho, 2015). Those studies connecting the Sunflower Movement to elections have assumed that young people had substantial effects on election outcome by voting (Hsieh, 2015; Subba, 2016). Indeed, the question of whether the movement benefitted young candidates has been neglected. This article focuses on the performance of young candidates in the post-Sunflower village and chief elections. If the energy of the movement carried into the election and the resources provided to candidates were helpful, then empirical evidence of these effects should be found in the 2014 village election results. For example, more young candidates were expected to run for office and to receive more votes than during the previous election.
Examining the performance of young candidates in the 2014 village elections is especially important owing to the context in which this election was embedded. Human beings are mortal. Age is one of many critical personal characteristics of candidates that serve as cognitive heuristics of competitiveness and quality (Lau and Redlawsk, 2001). People tend to match specific jobs with specific age groups (e.g. Perry and Bourhis, 1998; Perry et al., 1996). In politics, middle-aged candidates are usually preferred to those who are considered too young or too old (Armstrong and Graefe, 2011; Banducci et al., 2008): youth implies inexperience, while old age suggests inability.
The 2014 village elections in Taiwan provide an extreme but generalizable case study. In most democracies, young candidates are less competitive because of their lack of experience and their opponents’ incumbent advantage, and the young generations are always under-represented (Thompson and Singh, 2018). Such ageism happens so routinely that people even see the young candidate as less competitive in the experiment controlling for all other candidates’ social-demographic backgrounds (Berggren et al., 2010). However, previous studies assumed the ageism (rather than the candidate’s age) as a constant rather than a variable. The 2014 village elections in Taiwan offer a unique opportunity for researchers to explore what will happen when the constant is shaken by a large-scale social movement. Since the atmosphere and resources favored young candidates, they should have performed better than young candidates in previous elections, controlling for the existing electoral system (single-member districts and a national bipartisan system) and political culture. If young candidates did not perform better in this election than in previous ones, the existing system may discriminate against young candidates to such an extent that even a powerful social movement composed of young people and additional electoral resources failed to overcome.
Village chiefs, the Sunflower Movement and opposition parties
In Taiwan, the village chief is the elected representative at the lowest level of administration. The Japanese colonial government created the village chief as part of a system of social control (Yao, 2008a). After the KMT government retreated to Taiwan in 1949, the position was opened to electoral competition, but it was soon integrated into the party machine of electoral mobilization and clientelism in the context of military rule (Wang, 1997; Yao, 2008b). Although Taiwan started its democratization process in 1987, the majority of village chiefs were KMT members or tended to support the KMT. They did not necessarily join the KMT because the districts were small enough that they could connect to enough voters to win an election without using the party brand (Wang, 2015b).
Therefore, members of DPP, TSU and civil groups believed that village incumbents contributed to the KMT base, so if new, young candidates stemming from the Sunflower Movement defeated these village chiefs, they could uproot the influence of the KMT in local politics. To achieve this goal, financial and educational resources and even party nominations were provided to encourage young people to run in the 2014 village elections. Such plans included the DPP’s Democracy Grass, 2 the TSU’s Young People Running in Village Elections, 3 and civic groups such as Let’s Run in the Village Elections 4 and Youth Occupy Politics. 5 For example, in the Democracy Grass plan, the DPP helped young candidates create slogans, campaign videos and business cards; provided funding of 30,000 New Taiwan Dollars (about USD $1000); and provided campaign consulting services, campaign volunteers and local canvassers. 6 Moreover, the DPP did not nominate other candidates in districts in which young candidates had registered. The plan was proposed and led by Tsai Ing-wen, the DPP leader who won the 2016 presidential election.
The strategies implemented by the DPP and other groups drew the attention of both domestic and international media outlets. 7 The deadline to sign up for the village elections was 5 September 2014, which indicated that candidates had at least five months after the Sunflower Movement to make the decision to run for office and to prepare for the elections.
Many young candidates also ran in the local council elections. However, the cost of running in a council election is considerably higher than that of a village election given the size of the district and the competitiveness of the elections, and incumbent partisan candidates dominate most of these districts. Therefore, cooperation between young candidates and the existing parties was quite complex, which may blur the relationship between generational effects and election results. The relationship between the Sunflower Movement and the council elections was reported on several times, and its relationship to the village elections received some media exposure. For example, between 30 May 2014 and 30 November 2014 (election day), the terms ‘Sunflower Movement’ + ‘city council election’ appeared 47 times in the Apple Daily, the biggest newspaper in Taiwan. 8 During the same period, the terms ‘Sunflower Movement’ + ‘village chief election’ appeared 20 times, which was noticeable coverage. Therefore, it is reasonable to exploit the 2014 village and chief elections to examine whether social movements push members of the young generation to run in elections.
Exploring the influence of social movements in local elections
This article examines three different aspects to explore the performance of young candidates in the 2014 village elections: supply, demand and spillovers.
Supply side
First, if the Sunflower Movement raised the level of political participation among young people, more of them should have run in the upcoming elections. Winning in the elections is a direct way to punish the incumbent and to provide all voters with an alternative. If the Sunflower Movement was an effective treatment, we should observe more young candidates running in the post-Sunflower elections.
There are two measures to test this supply-side hypothesis. The simple test is to compare the number of young candidates between this and last election. Alternatively, it could be possible that the political environment encouraged more people from all generations to run in this election so the overall number of candidates will increase, but the number of young candidates increased more than other generations. If it is the case, the proportion of young candidates should be higher than the previous election.
H1-1: The number of young candidates in the village elections is higher in 2014 than in 2010
H1-2: The proportion of young candidates in the village elections is higher in 2014 than in 2010
Demand side
Second, the Sunflower Movement may have changed stereotypes of the young generation, or it might make the young generation a salient issue. For example, Madestam et al. (2013) found that a successful Tea Party protest (i.e. one not stopped by rainfall) increased public support for the Tea Party’s positions, and therefore increased the vote share of the Tea Party candidates in the upcoming election. Similarly, if the Sunflower Movement featured its young Taiwanese protesters and the young generation, the young candidates should be much more attractive in the election after the Sunflower Movement.
Moreover, the Sunflower Movement was an essential political opportunity for parties to attract new supporters. For example, Kitschelt (1988, 1993) analyzed the emergence of left-libertarian parties in Europe, and he argues that social movements provide motivation and attract educated and young voters. Although cleavages already existed and were made salient by the movement, it was the organized parties that exploited these cleavages and attracted voters in the elections after the social movement.
In this case, young candidates should have received more votes than in previous elections. Moreover, young candidates may have increased voter turnout if people perceived a high expected utility of young candidates winning the election.
H2-1: The vote share young candidates received is higher in 2014 than in 2010
H2-2: The number of chief positions won by young candidates is higher in 2014 than in 2010
H2-3: The turnout rate is higher in districts with young candidates than in those without
Spillover effects
Apart from affecting demand and supply, the emphasis on the young generation of the Sunflower Movement may have created a new issue dimension for voters (Deegan-Krause, 2007). Following the spatial theory of voting, an additional issue dimension may change the relative spatial distances among candidates and voters, which upends the status quo of local politics. The village chief is elected through a single-member district, so the entry of young candidates may reduce the vote shares of both incumbents and other challengers.
In local elections, incumbents usually enjoy considerable advantages, and the village chief elections in Taiwan are no exception. Analyzing the results of the 2010 and 2014 village chief elections, Wang (2015b) suggests that incumbents received, on average, 10% more votes than opposition candidates. In 2014, 70.2% of incumbents were re-elected.
Theories in behavioral psychology argue that people do not evaluate candidates in a vacuum; instead, voters’ evaluations of a candidate may be influenced by their perceptions of other candidates in the same election (e.g. Wang and Chen, 2018). When a new and young candidate enters the race, this new face primes voters in the district to think of the Sunflower Movement that happened earlier in the election year. It is also likely that a young candidate actively inserts the element of youth into his/her campaign framing and encourages voters to think of this characteristic.
Once a young candidate decided to step onto the battlefield, who would be influenced more? In Taiwan, previous electoral studies have yielded no consensus on the relationship between a candidate’s age and his/her competitiveness. Some studies suggest that Taiwanese voters prefer middle-aged candidates in some elections (Huang and Lin, 2007; Wang 2015b), while others fail to find such a correlation (Hsu and Lin, 2012; Sheng, 2008; Wang, 2015a).
Most village chief incumbents were KMT based or KMT friendly, and previous studies show that young voters show less support for KMT (e.g. Ho et al., 2015; Hsieh, 2016); thus, it is likely that young candidates and the new dimension of age in the post-Sunflower elections decrease the vote shares of challengers (mostly non-KMT or non-KMT-friendly) more than of incumbents. If a young candidate decides to run in a village and chief election instead of cooperating with the existing challenger, this decision would increase the difference between the incumbent and the challenger, making the incumbent less likely to step down.
H3-1: Adding one young candidate to a district increases the incumbent’s vote share
H3-2: Adding one young candidate to a district increases the likelihood of the incumbent being re-elected
Data and measures
The 2014 and 2010 village elections are the only two for which results are available on the website of the Central Election Commission. 9 Using the Sunflower Movement as a treatment, the hypothesis will be tested by comparing the results of the 2010 and 2014 elections. Data on each candidate’s age, gender, district, incumbency and vote share are available on the website, which are sufficient for hypothesis testing.
The hypotheses center on the term ‘young candidate’, but what is the definition of young? In this article, candidates younger than 40 are defined as young candidates. This definition was used at the time by the two major parties, KMT 10 and DPP, 11 to define their youth leagues and youth members. (After 2017, the threshold was changed to 35 by both parties.)
Before proceeding with hypothesis testing, brief descriptions of the 2014 and 2010 village elections can help provide the context and characteristics of the elections.
In the 2014 village elections, there were 7848 districts. Of the 14,137 candidates, 2633 (18.6%) were nominated by the KMT, 718 (5.1%) by the DPP and 45 by other minor parties; most candidates were non-partisan. However, as previously mentioned, for historical reasons most of these non-partisan candidates were pro-KMT. Overall, 6921 incumbents sought re-election, and 5513 were successfully re-elected. Therefore, the re-election rate was about 70.2%, which was not uncommon in local elections with single-member districts.
The 2010 village chief elections reflected similar patterns. In 2010, there were 7831 districts. The number of districts differs from that in 2014 due to minor district realignments. Of the 15,428 candidates, 3506 (22.7%) were from KMT, 583 (3.7%) from DPP and six from minor parties. Again, most candidates were non-partisan. However, due to the growth of many counties and cities (and the creation of the five special municipalities), data on incumbency in the 2010 election are incomplete. However, 74.5% of village chiefs were re-elected in Taipei City, and 71.5 % were re-elected in counties excluding the five special municipalities. It is worth noting that the DPP nominated fewer candidates in this election, a phenomenon that should be explored in future research.
Overall, the 2014 and 2010 village elections are similar in terms of many contextual factors. Therefore, a comparison of these two cycles can help estimate the influence of the Sunflower Movement that took place between them.
Performance of young candidates after the social movement 12
Supply side: more young candidates?
If the Sunflower Movement successfully raised the level of political participation of the young generation, and if the subsidies and resources provided by opposition parties and civic groups worked, then we should observe more young candidates in local elections.
In the 2014 village elections, there were 730 candidates under age 40 or 5.2% of the 14,137 candidates. The average age of all candidates was 56.8, with a median of 58. In comparison, in 2010, there were 901 young candidates or 5.8% of the 15,428 candidates. The mean and median candidate age was 55.1 and 56, respectively. To consider the influence of incumbency, 599 of 730 young candidates in 2014 were challengers, whereas in 2010, 797 of 901 candidates were challengers.
It is possible that the overall number of young candidates was smaller in 2014 but that the number nominated by opposition parties increased. 13 Testing such a hypothesis is not easy because the DPP’s plan included training for non-partisan candidates. If we only focus on young candidates nominated by the DPP, there were 70 candidates in 2014 (of 718, 9.8%) and 53 in 2010 (of 583, 9.1%). A Chi-square test shows no statistical difference between the two years (p = 0.55). Even though the DPP nominated more young candidates in 2014, the overall proportion was about the same as in 2010. Therefore, DPP nominations offer limited support for H1-1, but the evidence is far from sufficient to argue that the Sunflower Movement substantially affected the number of young people running in elections.
Since these two datasets are for the population instead of samples, statistical testing is unnecessary; comparing the means is enough for falsification. The evidence fails to support H1-1 and H1-2 – the number of young candidates did not increase after the Sunflower Movement, despite the resources provided and an atmosphere benefiting youth.
Demand side: greater support for young candidates?
First, let us focus on the vote share. Although the Sunflower Movement failed to encourage more young people to run for office and provide an alternative on the ballot, it is possible that voters were influenced by the movement and found existing young candidates much more attractive. Therefore, voters turn out and support young candidates in the post-Sunflower election.
In the 2014 election, young candidates received an average vote share of 42.2%, which is lower than the overall average of 55.5% (the percentage is larger than 50% due to many uncontested districts). In 2010, young candidates received 42.8% of the total votes. If we focus on districts with only an incumbent and a challenger (n = 3276), on average, the challenger received 44.5% of the total votes, but the vote share of young challengers was 43.4%.
Moreover, 291 of 730 (39.9%) young candidates won their elections in 2014, but in 2010 the proportion was 390 of 901 (43.1%). Overall, 3.7% of winners in the 2014 election were under 40; in 2010 the proportion was 5.0%.
Once again, comparing 2010 and 2014 election results reveals that voters did not show more support for young candidates after the Sunflower Movement. The election results fail to support H2-1 and H2–2.
If we focus on young candidates nominated by the DPP, they received 43.6% of the votes in 2014 and 41.8% in 2010 – a 1.8% difference. Regarding the chances of winning, young DPP candidates’ chances of winning were 48.8% in 2014 and 43.4% in 2010. However, this increased chance of winning was affected by the presence of young DPP incumbents. Of the 17 young DPP incumbents, 16 were re-elected in 2014. Of the 53 young DPP challengers, only 16 defeated the incumbent (32.07%). In comparison, the DPP nominated 49 young challengers in 2010, and 20 of them defeated the incumbent (40.8%). Once again, the results for the DPP provide limited, if any, support for the two hypotheses.
Second, we can consider turnout. Theoretically, young candidates provide an alternative for which more people are willing to come out and vote. By this logic, a young candidate is defined as a treatment of existing local political competition.
Figure 1 shows the average turnout rate under different scenarios, and the parentheses under each condition indicate the number of districts in the 2014 election. In this figure, the effect of adding one more young candidate to a district on turnout rate is limited. It is clear that the turnout rates in districts with one old and one young candidate are higher than in uncontested districts, but the effect is indiscernible in districts with two old candidates. In other words, competition spurs turnout, but the effect has nothing to do with the age of the candidates. Since the previous section clarifies that the number of young candidates did not increase, it also implies that the Sunflower Movement did not make elections much more competitive and had little effect on the turnout rate.
Turnout rate in districts with different candidate composition.
However, it is possible that there is an endogeneity problem when estimating the effect of young candidates on turnout. It is possible that the decision of a young candidate to run is related to the conditions of the district in previous elections. If so, the treatment of running as a young candidate is biased, and the estimation of the effect is therefore biased.
To deal with this problem, Rubin (1974) suggests propensity score matching to estimate the treatment effect by matching close cases in the dataset. In the first step, the treatment assignment is estimated by using a logit model in which the dependent variable is binary (treated or not) and independent variables include all covariates that could influence the assignment. The model is then used to generate the probability that each case is treated, and cases with the same probability of treatment but different conditions (treated or not) are paired. This step helps eliminate bias in the treatment assignment. Then all we need to do is compare the means of the treatment and control groups, and the difference is the estimated treatment effect.
For the case of the 2014 village elections in Taiwan, the treatment is adding one more candidate (either young or old) to the district. The covariates include the effective number of candidates in the 2010 election, whether the incumbent is seeking re-election and the number of the electorate in the district. The MatchIt package is used for estimation in R using the nearest neighbor method with a ratio of 1.
After propensity score matching, the pure effect on the turnout rate of an additional candidate in a district with a single candidate is 6.3% (p < 0.000, from 66.6% to 72.9%). In comparison, the effect of adding one more young candidate is 4.5% (p < 0.000, from 66.3% to 70.8%). For districts with two candidates, the effect on turnout of an additional candidate is non-significant (0.3%, p = 0.41, from 71.8% to 72.0%); the effect is the same for an additional young candidate (0.59%, p = 0.51, from 70.6% to 71.2%). Both point estimation and propensity score matching fail to support H2–3.
To summarize, young candidates did not perform better in village elections in 2014 than in 2010. They did not receive more votes after the Sunflower Movement. Moreover, the number of young chiefs dropped from 390 to 291 over those four years. In terms of increasing turnout, the effect of adding one more young candidate is indistinguishable from that of adding an older one. Overall, there is no evidence that the Sunflower Movement increased political participation of either candidates or voters in the 2014 village elections.
Spillover effects: disrupting the status quo?
In village elections in Taiwan, it is not uncommon for an incumbent to compete with the same challenger more than once. In the 7848 districts in the 2014 elections, 1247 (15.9%) incumbents faced the same opponent they had defeated 4 years ago. In 2014, the re-election rate was about 70.2%, which implies that challengers have a low chance of winning. However, many challengers were strong enough to fight more than once. For example, Tsai and Wang (2007) note that challengers in local politics create or become the leaders of community development associations in order to develop their ability to challenge incumbent village chiefs in the next election.
What would happen if a young and inexperienced candidate suddenly jumped into the race and became a third option? Traditionally, competition in single-member districts like those for village elections centers on two major candidates. They mobilize voters through personal networks, clientelism or party identity, all of which are relatively stable over time. A new face in the competition introduces at least one more issue dimension into the existing election context – age – which might change the relative distance between existing candidates and voters.
In the 2014 village elections, 49 (of 1247) districts fit this scenario. The incumbent and the challenger ran in both 2010 and 2014, but a young candidate joined the race in 2014. To estimate the effect of additional young candidates, propensity score matching is used. The control group includes all districts with an incumbent and a repeat challenger, and the treatment group is defined as all districts with an incumbent, a repeat challenger and a new young candidate. In the matching equation, the two independent variables are the number of effective candidates in the district in 2010 and the size of the electorate. The number of effective candidates is used because it captures the chance of winning and quality of both the incumbent and the repeat challenger, while the size of the electorate captures the socioeconomic development of the district. After matching by the nearest neighbor method, 49 treatment-control pairs are selected.
Figure 2 shows the effect of adding one more young candidate to the district on the vote share difference between the incumbent and the challenger. In the control group, which is shown in the figure as the red line with the 95% confidence interval, the vote share difference between the incumbent and the challenger in 2010 was, on average, 16.0%. In 2014, the difference narrowed to 8.1%; the incumbent received, on average, 54% of the vote share, while the challenger received 46%. The smaller gap may be attributed to the increased competitiveness of the challenger or an incumbent disadvantage. For example, even though voters are choosing between the same options, many voters could change their minds based on retrospective considerations.
Effect of adding a young candidate on the gap between incumbent and challenger.
In the treatment group, however, the difference between incumbent and challenger is 19.5% in 2010. In 2014, when a young candidate joined the race, the difference dropped sharply to 3.3%: the incumbent received 40.4% of the vote share, the challenger 37.1%, and the young candidate 22.5%, on average.
This analysis is informative for many aspects of elections. A new, young candidate will attract votes from both the incumbent and the repeat challenger, but, on average, the incumbent is more affected. Even though the young candidate has the lowest chance of winning, he/she can indirectly influence the relative strength of the incumbent and the challenger. In the spatial model, the young candidate represents a new reference point, which makes the incumbent seem more extreme or the challenger seem more moderate. Therefore, the evidence clearly rejects H3-1 – adding one young candidate in fact decreases the gap between incumbent and challenger.
A similar pattern can be found in Figure 3 for the change in re-election odds. In this figure, the dependent variable is the incumbent re-election rate in 2014. When competing against the same opponent without interference from other candidates (that is, in the control group), the winning rate in 2010 is 100% since they were incumbents in 2014. However, in 2014, the re-election rate is about 71.4%, which is close to the population mean (70.2%). When an additional young candidate joins the race, incumbent re-election rate drops precipitously to 45.0%. The evidence conflicts with H3-2. Therefore, adding a young candidate indirectly reduces the re-election rate. 14
Effect of adding a young candidate on the re-election rate of the incumbent.
Conclusion: the flower wilts?
Did a social movement featuring youth political participation encourage more young people to run in and help them win local elections? An analysis of the performance of young candidates in the 2014 village elections in Taiwan reveals both bad and good news. The bad news is that there is no evidence that more young people chose to run in local elections after the Sunflower Movement, nor did they become much more attractive to Taiwanese voters. After the 2014 elections, there were in fact fewer young chiefs than in 2010. The electoral resources provided by the DPP, TSU and other civil groups did not significantly affect this election.
The good news, however, is that once a young candidate decides to join the race, the evidence shows that he/she can indirectly change the results; the young candidate will not win, but he/she will help the challenger defeat the incumbent.
In a democratic system, the most common and legitimate form of changing policies and holding authorities accountable is through elections that can replace the incumbent. As discussed at the beginning of this paper, the 2014 village elections provide an extreme but generalizable case. Given the atmosphere of the social movement and the electoral resources provided to candidates, more young people should have chosen to run in the 2014 elections. However, the empirical evidence does not support this expectation.
This null finding suggests that the barriers to running in elections are high for Taiwanese young people given that the Sunflower Movement and additional resources failed to overcome them. Despite their ability to change election results, young people chose not to run. In Taiwan, people under 40 accounted for 38.4% of the adult population in 2014. 15 Yet, in villages, less than 5% of chiefs are young. The reasons for this pattern and discrimination against young candidates in Taiwan require further exploration; the Sunflower Movement seems to have not reduced this tendency.
A possible explanation may stem from the function of the village chief. Although village chiefs were believed to be relevant to the ruling KMT, their actual power was insufficient to change policy at the national and county levels. Therefore, even though opposition parties and non-governmental organizations (NGOs) called for young candidates to alter local politics by becoming village chiefs, young people may prefer to be councilors or even legislators. These calculations within the young generation will be studied in the future. Moreover, opposition parties may have wasted their resources and miscalculated in their campaign strategies in the 2014 local elections. Therefore, future work should examine whether more young candidates run in the 2016 legislative elections than in the 2012 elections. Nevertheless, it is still surprising that the Sunflower Movement had no direct effect on the decisions of young Taiwanese to run in the lowest-level elections.
Achen and Wang’s (2019) recent analysis of the generation gap in turnout in Taiwan also speaks to this article. After correcting for over-reporting, they found that turnout among the young voters dropped the most compared with other generations. Specifically, turnout rates among 20–29- and 30–39-year-olds were much lower in 2016 than in 2012 and 2008. Achen and Wang’s results provide additional evidence that this social movement did not encourage young citizens to engage in politics in traditional ways, as suggested in this article.
The third possible explanation of this result is the fluctuated and multi-dimensional support of the Sunflower Movement. Ho (2015) cited the TVBS polls conducted during the occupation, which showed that the 70% of Taiwanese people supported the occupation in the first few days (while 20% opposed the movement). But most of the supporters turned to non-responses after two weeks, and the level of support dropped from 70% to 26%. Meanwhile, the poll also showed that 70% of respondents agreed with one of the main goals of the occupation: reviewing the CSSTA on an article-by-article basis. Therefore, it could be possible that the support towards the Sunflower movement ebbed faster than the expectation of the opposition parties, so their strategies of encouraging young candidates were not as effective as they planned.
It is worth noting that this article does not distinguish between young candidates who were encouraged by the Sunflower Movement and those from political families. In Taiwan, it is common for the village chief’s son, daughter, wife or other relative to replace the incumbent. Therefore, both the incumbents and the young candidates are measured imprecisely and suffer from contamination. However, political family effects would have enhanced the performance of young candidates in the 2014 election, but improved performance was not observed. Therefore, I assume that the effects of political families do not influence the overall contribution of this article. However, future work should focus on the potential effects of political families on local elections and explore better measures of the incumbent.
Supplemental Material
clear_treatmentVR_1 – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, clear_treatmentVR_1 for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Supplemental Material
clear_treatmentWIN_1 – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, clear_treatmentWIN_1 for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Supplemental Material
create_appendix – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, create_appendix for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Supplemental Material
list2010_1 – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, list2010_1 for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Supplemental Material
list2014_1 – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, list2014_1 for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Supplemental Material
test_li_1 – Supplemental material for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan
Supplemental material, test_li_1 for Do Social Movements Encourage Young People to Run for Office? Evidence from the 2014 Sunflower Movement in Taiwan by Austin Horng-En Wang in Journal of Asian and African Studies
Footnotes
Acknowledgements
I am grateful to the editor for facilitating the review process, and for the constructive suggestions from the reviewers. I would also like to thank Fang-Yu Chen and Lu-huei Chen for the helpful comments on previous versions. All errors are on my own.
Author’s note
The preliminary version of this article was presented at the 2015 North American Taiwan Studies Association annual conference at Harvard University.
Funding
The author received no financial support for the research, authorship and/or publication of this article.
Notes
Author biography
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
