Some head starts are bigger than others: dynastic legacies and variation in candidate quality in Taiwan's local elections

Abstract

Candidates from established political families are more likely to win than their non-dynastic counterparts because their political inheritance often includes significant advantages in resources and reputation. Yet dynastic candidates also bring their own set of experiences and qualities to the game. Is it the case that their individual characteristics can explain their electoral success, or is their success due to their family legacy? Theoretically, how and when are political resources transferred to the new generation? We examine these questions by looking at non-incumbent candidates in city and county council elections in Taiwan from 2009 to 2014, drawing on unique data on politicians' backgrounds. The profile of dynastic candidates differs from that of non-dynastic candidates: they are younger, have less electoral experience and list experience in different sorts of organizations and jobs. However, these differences are not what drives their electoral success; indeed, they tend to win despite these qualities rather than because of them. Even after controlling for candidate quality, dynastic status remains a powerful predictor, conveying roughly a 20 percentage point increase in the probability of winning. This suggests that while dynastic candidates accrue some of their advantage long before they run for office, a large part of the inheritance is transferred during the campaign.

When 29-year-old Lin Ting-chun 林亭君 threw her hat in the ring for the 2014 Taipei City Council election, her uncle Lin Chin-chang 林晉章 accompanied her to the election commission office to formally register her candidacy. At the time, he was completing his sixth consecutive term in the council and had developed a reputation as an elder statesman, leading exchanges with local councils in other countries. The youngest of the nominees of the Kuomintang (KMT) in the capital city that year, the niece's main qualification was her 4 years on her uncle's staff, including as director of his constituent service centre; punning on her name, her campaign slogan promised continued service (您挺君 君挺您).¹ The senior Lin assured his constituents that if they elected her, they could count on his ongoing presence – he would bring his 25 years of experience to bear in assisting her and they ‘need not worry that either oversight of city government or quality of service would suffer’.² Recruited by a family elder seeking a successor, Ting-chun truly owed her candidacy to her uncle and ran heavily in his shadow. She eventually won a seat, coming in sixth in the eight-seat district.³

Though he also followed in the footsteps of a close family member, the path Chiang Chao-kuo 江肇國 took to a city council seat was substantially different. Chiang's father, Chiang Cheng-chi 江正吉, served four terms in the Taichung City Council before losing his seat in 2002. Yet Chao-kuo did not run for any office until 2014, when, at age 37, he took aim at the Taichung council. His father had thus been out of office for a full 12 years; moreover, while in the council he had been part of the KMT, whereas Chao-kuo ran with the Democratic Progressive Party's (DPP) nomination.⁴ Prior to running, the son went out and systematically built the resume of a good politician. He earned a masters at the renowned National Chengchi University; worked as a legislative aide at the national level; and served on mayoral and vice-presidential campaigns. He even garnered headlines for marrying a woman who served as a bodyguard for Tsai Ing-wen in the 2012 presidential election.

Studies of political dynasties often start with a story much like Lin Ting-chun's (Smith, 2012, 2018; Ockey, 2015; Bohlken, 2016; Chauchard, 2016). In such accounts, the senior politician decides to hand off his seat to a young family member, who has little or no political experience and indeed is barely out of school. Sometimes, this successor isn't even particularly interested in politics. To the degree that the successor has done anything relevant, it is usually working in the predecessor's office or campaign or in some other junior position arranged by the predecessor. Since the legacy candidate is a political ingénue, the burden of running the campaign falls entirely on the previous generation. Nevertheless, since the predecessor has years of experience, an ample war chest, an extensive political network and a good reputation with voters, the legacy candidate easily wins the race.

This ‘empty shell’ caricature is an inadequate depiction of dynastic politics, however.⁵ While some legacy candidates resemble this image, many others differ in important ways. For example, Chiang Chao-kuo came to his first city council election already having practiced politics at multiple levels. Chiang clearly owed a political debt to his father's connections and financial resources, and no doubt began learning the political trade at a young age. Yet much of his viability as a first-time candidate came from his own long-term efforts and career-building. Some legacy candidates come to the campaign with formidable networks from work with corporatist bodies, NGOs, clubs or businesses. In some cases, they sharpen their credentials by pursuing advanced degrees. Finally, not all legacy candidates are young; some do not plunge into politics until middle age, having built career experience in domains outside of politics. One goal of this paper is to paint a fuller portrait of dynastic candidates, highlighting their considerable diversity.

It is clear from a large body of literature that dynastic candidates are more likely to win elections than their non-dynastic counterparts. However, because the diversity of dynastic politicians has been largely overlooked, it is not clear whether this success stems directly or indirectly from their political inheritance. The literature suggests that dynastic candidates benefit by inheriting various types of political resources. However, they might also build up their own store of networks and abilities years before entering an election race by engaging in political activities and building relevant experience through their own efforts. Do dynastic politicians do better because their predecessors directly transfer resources to them, or do they do better because they are able to build up their own resources before entering politics? More bluntly, to what extent can candidate quality explain the success of dynastic candidates?

Employing a carefully assembled, one-of-a-kind dataset, we examine the electoral success of non-incumbent candidates in city and county council elections in Taiwan from 2009 to 2014 and find that the experiences that dynastic and non-dynastic candidates bring to the election vary and that, in general, candidate quality has an impact on electoral success. However, many of the specific qualities that legacy candidates are more likely than others to have are not associated with winning. Even controlling for candidate quality, dynastic candidates are significantly more likely to win than their non-dynastic counterparts. This is powerful evidence for the argument that dynastic predecessors directly transfer important assets – whether in the form of political know-how, campaign funds and organization, associational ties, name recognition or other resources – to their successors.

All told, these results are somewhat troubling for democracy. There are quite a few Chiang Chao-kuos, whose entrée into the heart of Taiwan's electoral politics via a city or county council arguably owes as much to their own efforts to cultivate candidate strength as to a received inheritance. Yet cases like those of Lin Ting-chun, in which resources more or less directly transferred from a previously elected family member figure heavily, are more the norm than the exception. Such legacies alone confer an immense advantage amid the pool of newcomers competing for scarce seats without prior council experience. To the extent that these directly transferred resources determine election results, it is harder for non-dynastic challengers to build comparable strength and break into the political realm. Happily, it is far from impossible to do so, however; in fact a substantial majority of newcomers who won council seats between 2009 and 2014 (71%) had no family political legacy.

1. Why are dynastic candidates more successful?

It is not an uncommon observation that dynastic politicians do well in democratic politics. Scholars have documented relatively high numbers of dynastic politicians in a variety of countries, including Japan (Smith, 2012), the Philippines (Mendoza et al., 2012), Brazil (Marques, 2018), Ireland (Smith and Martin, 2017), India (Chandra, 2016), the UK (Van Coppenolle, 2014), Thailand (Ockey, 2015), Iceland (Smith, 2018: 5) and Taiwan (Batto, 2018a), just to name a few. These findings have led a number of scholars to some variant of Dal Bo et al.'s concise yet vivid conclusion that ‘power begets power’ (2009: 115; see also Querubin, 2016; Rossi, 2017; Smith and Martin, 2017). But how does power perpetuate itself?

The literature suggests several reasons that dynastic politicians might have an electoral advantage over non-dynastic politicians. Many of these take the form of resources that can be directly acquired from a family predecessor and deployed in campaigns. Dynastic politicians inherit superior financial, organizational, networking and institutional knowledge advantages from previous generations, and these resources can be used to outcompete non-dynastic rivals in struggles for nominations, for votes in general elections and in power struggles once in office.

Political finance works differently in every country, so it is difficult to empirically establish that dynastic politicians do, in fact, come to elections with more financial resources.⁶ Nevertheless, there are numerous anecdotes of well-financed legacy candidates, and popular media reports generally take it as obvious that second- or third-generation politicians will have ample financial resources. Being well-financed does not necessarily imply that dynastic families are all deeply engaged in systemic corruption. A less drastic way of thinking about it is to start from the notion that some sort of financial foundation is necessary to support almost every political career. Successful politicians have generally figured out some way to reliably and sustainably meet their financial needs. Some will settle on corrupt strategies, but many will be able to rely entirely on legal and aboveboard methods to raise the necessary funds. At any rate, when they decide to pass their power on to the next generation, they can transfer that strategy – as well as some of their amassed financial resources – to their designated successor.

In some cases, money can be directly turned into votes via vote-buying. More commonly, money is used more indirectly as a means of building up a formidable campaign structure. However, money alone is usually not sufficient. A healthy campaign organization requires campaign expertise, local knowledge, management skills, a wide array of personal contacts and an ability to communicate with large numbers of ordinary voters. Postwar Japanese koenkai are a classic example of such an organization. Japanese politicians built up massive support groups with complex organizational structures and often tens of thousands of ordinary members. The politicians had regularized contacts with their voters through these mini-bureaucracies, and these relationships yielded reliable support in elections. When the politician retired, the organization could simply be transferred to a designated successor, often a son (Curtis, 1988; Ishibashi and Reed, 1992). While Japanese koenkai are a rather extreme example, the logic of dynastic inheritance is similar in many contexts. The established politician builds up an organization over several years which can then be transferred to a child, spouse or other family member.

If local organizations can be understood as human capital at the local level, dynastic politicians may also have an advantage in networking at the elite level. Over the course of a long career, most politicians build up a network of important relationships. These may include party leaders, government executives, business leaders and others. If these resources are transferred to the next generation, dynastic successors may enjoy significant advantages in areas such as securing nominations, legislative bargaining and intra-party power struggles. To put it in more concrete terms, a top party leader might not have time for a telephone call from a relatively anonymous junior politician, but if that junior politician is the daughter of an old friend and ally, the party leader might decide to take the call. There is evidence from Japan that dynastic legislators are able to secure more distributive resources for their local area, which is consistent with the idea that they have systematic advantages in legislative bargaining (Asako et al., 2015).

Political knowledge is a fourth resource that dynastic politicians may enjoy. It is surely helpful to have an experienced, trustworthy advisor who knows how the system actually works. Aspiring politicians who follow in the footsteps of a family legacy generally have a go-to consigliere who knows how to operate the levers of power, whether that is in constructing a campaign message, dealing with specific party leaders or making demands on the bureaucracy. It cannot hurt to have someone around who has been through a few battles (Parker, 1996: 88).

Campaign resources are not the only kind of advantage that dynastic successors receive. They also may benefit from factors that have been pre-implanted in the minds of voters: higher name recognition, an established brand, perhaps even a popular preference for dynastic politicians. The impact of such intangibles remains a matter of debate. A recognizable surname may bring substantial benefits (Ravanilla et al., 2022), though overly common surnames might negate that effect. For example, since over 10% of the Taiwanese population is surnamed Chen, it would be unrealistic to assume that Taiwanese voters might connect one Chen to another family member named Chen. On the other hand, a Filipino voter seeing a famous dynastic name like Roxas or Aquino might take notice, since there are relatively few people with those surnames who are not part of the famous political clans. However, name recognition only helps a candidate stand out from the crowd; it does not necessarily make voters prefer that candidate. In fact, evidence from grassroots Filipino elections suggests that more recognizable surnames are not more likely to be elected (Cruz et al., 2017).

A more nuanced argument suggests that dynastic politicians enjoy a strong brand. Successful politicians often establish a reputation for acting in certain ways or pursuing particular policies, and a dynastic successor may be able to inherit this brand. This may help dynastic politicians differentiate themselves with a clear, credible appeal. Dynastic politicians may thus be able to bypass one of the first and highest hurdles facing any junior politician: communicating to voters who you are and what you stand for. The content of the brand could be quite specific. For example, it might be a reputation for fighting for a specific economic, environmental or cultural policy. Alternatively, it could be a vague impression that the family has performed well in office. For example, an Indian dynastic politician claiming a heritage of good governance explained ‘This family system runs because of credibility … Why do people want to buy a Mercedes car? Or a BMW car? Because they know the credibility of that car. You come out with a new car that nobody knows, nobody will buy it’ (Mandhana, 2014, quoted in Chandra, 2016: 30).⁷ While dynastic candidates may carry a distasteful odour of privilege for some voters, large numbers of voters in a variety of settings around the world in fact seem to look favourably on dynastic status.⁸

In this study, we do not measure these campaign- or brand-related electoral resources; we simply posit that they are likely useful for winning elections and that they directly accrue to dynastic candidates. In temporal terms, the transfer of these resources occurs proximately, just when the successor runs for office; they take effect during the campaign itself. However, there are a variety of ways to build up strength as an aspiring candidate; inheritance is but one. With or without a family legacy, a person might spend years putting together an extensive resume, constructing networks, amassing wealth, working for a political cause, garnering fame, becoming an effective public speaker and so on before she decides to jump into the electoral arena. Almost no one comes to an election as a blank slate. Almost everyone, both dynastic and non-dynastic, has some sort of relevant experience and has amassed some amount of relevant resources or reputation.

The literature on political dynasties largely ignores variation in candidate quality, though there are a few exceptions. Some Japanese dynastic families are wealthier than others (Asako et al., 2015), and dynastic Filipino legislators are wealthier than non-dynastic legislators (Mendoza et al., 2012). Neither study examines whether these differences in wealth affect electoral outcomes. A study on Norwegian elections collects data on candidates' occupations, but this does not appear to be systematically related to dynastic status or electoral success (Fiva and Smith, 2018). A few studies look at previous political positions and generally find that incumbency, length of tenure and holding other electoral offices are associated with electoral success (Smith and Reed, 2013; Asako et al., 2015; Fiva and Smith, 2018). While these studies note that dynastic candidates do not look identical to non-dynastic candidates or to each other, they do not adequately portray the full diversity within the set of dynastic candidates. None, for example, examines each candidate's politically relevant set of experiences or social networks. As such, the effect of candidate quality on the electoral success of dynastic candidates remains an open question.

A focus on candidate quality forces a reconsideration of how the dynastic heritage is transferred from one generation to the next. The default assumption is that the younger family members are largely empty shells, with little or no personal experience or political heft of their own. The dynastic predecessor directly channels resources to them, spending money, conveying an organizational apparatus, devising strategies, mobilizing networks and so on. In the empty-shell caricature, the candidate's personal qualities are meaningless. If this scenario adequately reflects reality, there should be no correlation between candidate quality and electoral success for legacy candidates. Their dynastic status alone should give them a substantial edge.

However, what if dynastic candidates are fully developed politicians rather than empty shells? It could be that the dynastic advantage is transferred long before the legacy politician enters the race. As the child of a prominent politician, the future politician might be favoured by teachers or encouraged to take leadership roles. As a young adult, the future politician might be able to enter doors that are closed to other people, landing jobs that train them for a future political career or meeting people that can help them advance that career. Powerful people might strategically place bets on them as potential future power brokers. By the time the aspiring junior politician decides to venture into the electoral arena, he or she might have leveraged years of preferential treatment to construct an impressive resume, formidable financial and organizational resources, and even begun building his or her reputation. The junior politician's advantages are already considerable enough that there is no need to consider resources transferred from the senior politician during the election campaign.

In this scenario, after considering candidate quality, dynastic status may not be correlated with electoral success. The dynastic advantage is fully embedded within the candidate's education, party nomination, work experience, previous electoral experience, experience with other important organizations and so on.

A third alternative lies somewhere between these two extremes, blending these two stories. On the one hand, due to their family status, future dynastic candidates have opportunities unavailable to other aspiring politicians. This might enable them to build up some electorally useful resources. On the other hand, because they can also directly tap into their family's amassed resources, they do not need to wait to become fully mature politicians before wading into the electoral arena. They can bypass the slow, arduous task of building up a thick resume because senior party leaders or other power brokers are willing to bet on them and voters are willing to support them for important offices right away based on their family ties. Each dynastic politician will have a unique blend of these two ingredients, depending on their willingness and ability to take advantage of the opportunities opened up to them and the magnitude of the resources the previous generation is able to directly transfer.

In this third scenario, we should see evidence that both dynastic status and candidate quality affect electoral outcomes. The dynastic advantage may be partially embedded in the candidate's resume, but it still has an independent effect.

2. Studying dynastic politics in Taiwan's city and county councils

In this paper, we scrutinize dynastic politics in Taiwan's city and county councils.⁹ As we explain below, this is an appropriate and fertile setting in which to study these questions. In comparative context, Taiwan has a fairly high proportion of dynastic politicians. Unlike many other studies that focus only on national-level politics, we look at the highest level of local politics, where the stakes are high enough for regular party politics to apply and there are ample cases to analyse quantitatively. Not least importantly, we were able to draw on a unique data source. Here, we explain the rationale in more detail for this particular empirical focus – the why – and the methods we have used – the how.

The twenty-two city and county councils constitute the vital middle level of Taiwan's politics. In substantive terms, they play a significant role in democratic governance. They pass ordinances, amend and approve the mayor's budget, and exercise oversight over local authorities, including through periodic interpellation sessions in which mayors and other officials face questions and account for their actions. They also act as a much-used channel between citizenry and local government. Taiwan's citizens regularly contact one of the island's approximately 900 councillors 議員 to request help on issues of many kinds, from roads and infrastructure to welfare benefits. Typically, a councillor figures prominently in local society, often making multiple invited appearances every weekend at weddings, funerals, festivals and other gatherings. Through reserved districts and a quota mechanism, councils are also sites for enhancing representation for indigenous citizens and women, respectively (on gender quotas, see Huang, 2015; Batto, 2019). City and county councillors have significant stature, figuring prominently in campaign appearances of candidates for the legislature and sometimes serving on leadership committees of the two major political parties. Those in large cities commonly make the national news with their press conferences or scandals, and some appear frequently on Taiwan's many political talk shows.

Most critically, these councils can be thought of as both coveted destinations and vital way-stations in the career trajectories of many politicians. Attaining a seat in a city and county council is a considerable accomplishment in itself. In counties, county council seats are more desirable than town council seats and an important stepping stone to town mayor positions. Everywhere, but especially in direct municipalities, city and county councils tower above the lowest elected positions, the neighbourhood heads (Read, 2012, 2020). For those with even higher political ambitions, particularly those aiming to become a national legislator, city mayor or county magistrate, the council is often an essential platform from which to build experience and resources for further ascent (Batto, 2018b). Given all this, councils make an important place to study dynastic politics. When political families exert influence and attempt to insert one of their offspring into elected offices, this level is often the target of their efforts. Moreover, far from being merely of academic interest, dynastic politics at the council level is keenly observed and discussed by the public. Much newspaper and magazine coverage has addressed the significant presence of legacy candidates, often adopting a critical tone towards ‘hereditary politics’ 世襲政治 and ‘second-generation politicians’ 政二代.¹⁰

Because there are large numbers of council candidates, elections at this level can provide for robust quantitative analysis of dynastic politics. For this paper, we built a unique dataset comprising two cycles of city and county council elections: 2009–10 and 2014.¹¹ A total of 3,087 candidates competed in elections in 20 localities¹² across these two cycles.¹³ Roughly half of these candidates were incumbents seeking re-election. Theoretically, the connection between dynastic status and electoral success is fairly tenuous for incumbents, who have already been in office for a period of time – sometimes a very long period – and may no longer depend as much (or at all) on inherited resources. As a result, we focus on the 1,581 non-incumbents, for whom the dynastic inheritance should be much more crucial to electoral success. We also do not include 207 individuals who were not running as incumbents but had previously held office at the city or county council level, on the presumption that they enjoyed at least some advantages of incumbency; in fact, results show little substantive difference with or without these individuals. The sample that we analyse in what follows thus numbers 1,374 candidates running for a council seat without ever having won one.

In this paper, we define dynastic candidates as those with a relative who has previously won an elected government office at any level since the Republic of China (ROC) assumed control of Taiwan in 1945, from president to neighbourhood head. We employed two sources in identifying legacy candidates. The first involved a systematic search of articles in the Lianhe Bao (United Daily News), Zhongguo Shibao (China Times) and Ziyou Shibao (Liberty Times). Typically, when new candidates first run for office, the local sections of these national newspapers run a story introducing them, mentioning family predecessors as well as other information such as previous offices and factional ties. Such stories thus provide a reasonably complete source of data on dynastic connections, although they are subject to certain forms of bias, such as variation in the quality of reporting and depth of coverage from place to place. This definition of dynastic candidates as well as the newspaper sources and their strengths and weaknesses are discussed in more detail elsewhere (Batto, 2018a: 495–497).

We augmented this source with data from political almanacs on the 2009 and 2010 elections compiled by researchers at a public opinion research firm (Hung, 2010; Liu, 2013). Those researchers visited each of Taiwan's cities and counties and interviewed local insiders to collect information on all candidates in council and other races, often including notes on family connections as well as factional affiliations, bases of support and more. Though the research firm has close ties to the DPP, the books are neutral in tone without obvious political biases, although they seem to contain more detail on some localities than others. These almanacs were a one-time project, not continued in later election cycles.¹⁴ Together, the newspaper reports and the almanacs identified 572 legacy candidates. Of these, 257 had never held a council seat.

Apart from data on a candidate's dynastic connections, we also wanted to learn more about his or her previous experiences and group affiliations, taking these as potential sources or markers of candidate quality. For this, we drew on the official election bulletins 選舉公報 that are published by local election commissions and circulated to voters' homes prior to each election.¹⁵ In addition to a candidate's photograph, name, ballot number, party nomination and other basic information, these bulletins convey a summary of the candidate's ‘experience’ 經歷. These mini-CVs list both work experiences and group affiliations. The associations mentioned span a wide range from lineage organizations to temple boards to alumni, sports and social movement groups; they include civil society organizations as well as state-linked bodies such as volunteer police and military reserve groups. As these are self-reported by the candidate, we must see them as strategic texts aimed at conveying a positive impression and emphasizing links to potential bases of support. As with resumes generally, there is a degree of ambiguity and probably of overstatement. When someone reports he was a board member of the Yilan County Wandefu Association for Promoting the Rights and Interests of the Disadvantaged 宜蘭縣萬得福弱勢族群權益發展協會 or she was an advisor to the Kaohsiung City Shude Women's Association 高雄市淑德婦女協進會, that could indicate a longstanding and deep commitment or perhaps a more ephemeral or honorary tie. Given the wide circulation of the bulletins, however, we assume that candidates generally do not simply make up affiliations out of nothing. Research assistants copied the full text of these statements of experience into a spreadsheet for later coding by word matching.

Candidates reported a very wide range of positions and affiliations in their experience statements. To make sense of them, we created a set of variables classifying individual elements within the statements into general categories. For example, some candidates mentioned belonging to a business-oriented club, such as the Lions Club 獅子會, Rotarians 扶輪社, Jaycees 青商會 or chamber of commerce 商業會 / 商業總會. Each candidate whose experience statement contained any of these keywords was coded ‘1’ in a variable for civic and business associations. Table 1 shows all 10 such variables whose possible connection to electoral success we explored in our analysis. In that table, the third column gives examples of some of the search strings that were used to identify candidates to whom the category applied. Depending on the content of his or her experience statement, a given candidate could belong to one, two or several such categories, or to none at all.

Table 1. Variables coded from candidate ‘experience’ statements

For basic information on candidates, including sex, age, education, party, as well as election outcomes, we use data from the Central Election Commission. We also collected further information on candidates' background. This includes their incumbency status¹⁶ as well as data on prior offices held. Finally, we supplement this quantitative data with interviews conducted with local council members and political insiders.

3. Profile of dynastic candidates

What do aspiring candidates from political families look like as a group, in comparison to their non-legacy competitors? And what kind of variation do we see among dynastic candidates? Our data allow us to present a precise descriptive profile of the members of this much-discussed political category. Here as elsewhere, we focus on non-incumbents, ignoring all candidates running to defend a council seat, as well as those with earlier experience as a council member.

To begin with, as Table 2 shows, legacy candidates tended to be much younger than non-legacy candidates – more than 7 years younger on average.¹⁷ Nine per cent of legacy candidates were under the age of 30 (vs just 3% of non-legacies) and 47% were under 40 (vs 17% of non-legacies). As the case of the youthful Lin Ting-chun illustrated, dynastic advantages can empower newcomers to attempt to vault into councils at an age where few others would so dare. At the same time, we see that by no means do all legacy candidates fit this image. Their median age was 41, and about a quarter were in their 50s or older.

Table 2. Summary statistics for non-incumbent candidates

A second striking feature of the pool of legacy candidates is that it is relatively balanced in terms of gender. A far greater proportion of legacy candidates were women (44%) compared to the proportion among non-legacy candidates (22%).¹⁸ One plausible explanation might run as follows: to reach the point of being ready to make a credible run for city or county council as a political novice, non-legacy candidates have to accumulate many years of experience through institutions and roles that may exert a gendered selection or bias, filtering some women out. And as women grow older, start families and assume disproportionate burdens of caring for children, the exertions of a political campaign may seem more and more daunting. Whereas relatives of those who have previously held office do not need to navigate through as much of that path; they have early access to the family's political resources whether they are men or women, or at least with less regard to gender. Additionally, gender quotas may play a role (Batto, 2019). Political families with intimate knowledge of a district's political dynamics may strategically select female members to run in circumstances where they stand to benefit from the gender quotas. This was so in the case of Lin Ting-chun; her district lacked any women incumbents running in 2014 but had two quota slots, thus presenting an inviting target for female contenders – though in the end the quota was not invoked because of strong performances by Lin and another woman candidate.¹⁹

A third way in which legacy candidates stood out is their relative lack of prior experience in elected office. Among candidates without dynastic predecessors, 43% had won office before. The most common such office is town council, and others served as town mayors, or less commonly neighbourhood heads. By contrast, only 21% of dynastic contenders had previously held any elected position.

Legacy candidates were, however, much more likely to have served as political aides and assistants – often working for their own senior kin. Fifty-two per cent of dynastic successors listed experience in such positions as assistants 助理, campaign supporters 助選員, chiefs of staff 辦公室主任, secretaries 秘書 and working in constituent service centres 服務處. Among other candidates, the corresponding figure was a mere 26%. Here again, the case of Lin Ting-chun exemplifies the dynastic ‘neophyte’ type; she had never run for, let alone won an elected office before, and indeed seems not even to have envisioned politics as her calling. Her several years of fielding constituent calls for her uncle compensated for this and also passed on, to some degree, certain advantages of his long incumbency, such as familiarity with voters in the district and a sense of how a politician engages with them and cultivates their support. Chiang Chao-kuo, meanwhile, reminds us that other legacy candidates take a different road. His experience statement listed his five prior positions as assistant, secretary, chief of staff and such, but none of them involved working for his father, and all were in national roles for politicians of a party (the DPP) other than the one his father belonged to while serving in the council.

In addition to (or perhaps because of) their family ties, dynastic candidates had somewhat more formal education than their non-dynastic counterparts. More of them had at least a vocational college degree (68% of legacies, 61% of non-legacies). But almost the same percentage held advanced degrees: 21% of legacies, and 20% of non-legacies. Certainly, not all dynastic candidates were like Lin and Chiang, with 4-year college or (in his case) masters degrees. Thirty-two per cent had schooling only to the high school level or less.

With regard to other forms of experience, legacy and non-legacy candidates differed only in a few observable ways. Legacies reported somewhat fewer associational memberships and other types of experience overall. They were a bit less likely to spend time joining groups based on family, lineage or native-place ties, and grassroots volunteer groups such as neighbourhood watch patrols, volunteer police or volunteer firemen. They also reported less involvement in organizations committed to social change, environmental protection or support for disadvantaged populations (11% among legacies, as opposed to 19% among non-legacies). They had slightly higher participation in business and sports groups.

Do legacy candidates ‘inherit’ associational membership? In Lin's case, of the five associations listed in her experience statement, one (managing director of the Lin Family Lineage Association of Zhongshan District) seems likely to derive from her uncle, who had served on the board of that group. Other items are more ambiguous. Leadership in the Taipei City Roller Skating and Ice Skating Association may represent her own interests and initiative, though her uncle had also mentioned a city-wide skating group in his 2010 statement. She was involved in two women's associations, which in Taiwan are usually not officially affiliated with the KMT but are among its array of umbrella associations. She might have joined these groups due to personal interest in women's issues, or her uncle might have arranged the positions through his party connections. Similarly, she also led a local community development group, and it is not clear whether this represents her own initiative or her family's connections. Chiang's profile in his 2014 run featured a diverse set of affiliations, including a civil defence team, an organization for the elderly, an animal protection group and a local chapter of the Lions Club. None of these groups overlapped with those his father listed in his last city council run. From these specific examples and from the broader patterns in the data, we see that despite their inherited advantage, dynastic candidates still cultivate ties with associations the way other candidates do (although less so among lineage, grassroots volunteering and social change groups). They are less diligent in accumulating markers of community participation, but not dramatically so. Yet they are able to do these things at a younger age than their non-legacy counterparts; as well, some of these associational ties may be part of what they inherit from their dynastic predecessor.

What about party affiliations? As Table 2 shows, 28% of legacy candidates ran with KMT nominations, and 20% with DPP nominations. Yet the corresponding figures among non-legacy candidates were similar (25 and 15%, respectively). Dynastic politics was thus not particularly associated with either of the two major parties, nor was either relatively free of it. Very few dynastic successors chose to run under the banner of a smaller party. Most established politicians are affiliated with one of the two big parties or are independents, and most dynastic successors maintain the same party affiliation. More strategically, smaller parties have a fairly dismal record in lower-level elections, so established politicians seeking to elevate a junior family member naturally eschew them as bad bets. In local elections, KMT and DPP nominees are almost always seen as credible, while small party nominees are usually seen as long shots. Almost half of legacy candidates ran as independents, which was also true of non-legacies.

To sum up: because of their family ties, in many cases the dynastic candidates have taken a big shortcut, or a different path entirely, to get to the point where they are able to make a run in a challenging race for city or county council. The much larger proportion of women among legacy candidates compared with non-legacy candidates reinforces the fact that they have been recruited, as it were, through different pathways. Legacy candidates run at a significantly younger age, and they are much more likely to leap into the ordeal of a council race without having previously won an elected office. Their family political capital makes up for the advantages and resources that age and prior office would provide. This capital does not mean they can forgo all the ways that ordinary candidates build or signal support from groups in society. Like other candidates, they pursue associational ties – or perhaps, in some cases, inherit them. Despite all these general tendencies, legacy candidates are by no means all alike. They exhibit substantial variation in factors that can matter in elections, including age, prior experience, party nominations and associational affiliations.

4. The dynastic advantage

Do dynastic candidates enjoy an advantage over other non-incumbents in elections for city and county councils, taking all things into account? If so, how large is that advantage? Our data allow us to provide reasonably confident answers to those questions.

Winning a council seat without the advantage of incumbency is no mean feat. All told, out of 1,374 candidates competing in council elections from 2009 to 2014 without previously having held a council seat, only 487 succeeded (35.4%). Dynastic candidates fared much better than those without family connections, however. Well over half of the legacy candidates (54.9%) won, compared to less than a third (31.0%) of non-legacy contenders. But several observable factors are correlated with victory. Through logistic regression analysis, we isolate the overall advantage deriving specifically from dynastic status.

Table 3 shows the results of models estimated on our dataset in which the outcome variable is whether or not a candidate won. The baseline simple model incorporates only the distinction between county and city council elections, a variable for gender,²⁰ and a candidate's dynastic (vs non-dynastic) status. The county variable accounts for the fact that it is easier for newcomers to win a seat in county councils than in city councils, at least in part due to a more favourable ratio of candidates to seats (in counties, 43% of non-incumbents won a seat, as opposed to just 28% of non-incumbents running in cities).

Table 3. Logistic regression models (DV: winning)

*P < 0.05, **P < 0.01, ***P < 0.001.

The full model also includes a range of variables that potentially relate to candidate quality. To begin with, this model shows – not surprisingly – that nominees of the two major parties, the KMT and DPP, have much better prospects of obtaining office than independents or those affiliated with smaller parties.²¹

The variable for age captures the fact that, in general terms, candidates' performance declines as they get older and older past their 30s. Strikingly, though, a particularly large gap in win rates appeared between dynastics and non-dynastics in the youngest age category, as shown in Figure 1. We saw earlier that the former often enter council races at much younger ages than the latter. What we see here is that those legacy candidates who run while still in their 20s dramatically outperform their non-legacy counterparts.²²

Figure 1. Win rates among non-incumbents, by dynastic status and age

Note: Number of candidates who had never held a council seat: 1,374. Among this group, no dynastic candidate was older than 66.

As explained earlier, among our goals in this inquiry was to understand the relationship between resources derived from various sources, including candidates' past experiences, and success at the ballot box. Education might be thought of as such an electoral resource; after all, candidates often make much of their academic pedigree, and the election bulletins prominently convey the highest level of schooling attained by each contender. Although education has a modest correlation with winning a seat, this factor does not hold up as significant in the regression analysis.

Another kind of experience matters greatly, however: those who previously held an elected office have a big leg up in races for city and county councils. As we saw in the last section, this often means having served in, for instance, a township council. In such cases, the candidate can draw on organizational resources, networks, name recognition and more from at least one previous election cycle and from a period of time in office. As we also saw, the legacy candidates more often than not did not have this factor working in their favour.

Our analysis turned up certain other aspects of candidates' past experience and organizational networks that affect their propensity to win. For example, contenders who reported an affiliation with what we call a corporatist group (specifically, a farmers association 農會 or fisherpersons association 漁會) had somewhat higher chances of success. This is true even for candidates in cities as well as those in counties, reflecting the ongoing relevance of such groups in local society, and the financial resources such as property and credit unions that these associations have historically controlled. Those who had served in what we call a grassroots volunteer capacity, including community development associations 社區發展協會 as well as neighbourhood watch and volunteer police groups, had a slight edge as well. In counties (but not cities), affiliation with sports and recreational groups (a category including everything from softball to dance to martial arts) also marginally improved a candidate's chances.²³

Even after accounting for the various candidate-quality factors that boost a contender's chances, the dynastic advantage stands out unmistakably in the model. This advantage is most easily expressed in the form of predicted probabilities of victory for a dynastic candidate as opposed to one who is alike in other respects but lacks the benefit of a family political legacy.

Figure 2 conveys how the dynastic advantage plays out for candidates with different qualities in an urban or a rural context. Each graph presents a point prediction for a non-dynastic and a dynastic candidate's chance of winning the election, respectively. (Bars represent 95% confidence intervals around the predicted values.) The vertical gap between the two points shows the dynastic advantage in that particular case. An independent (or third-party nominee) running without the benefit of prior office in a city council election, for instance, has just a 10% chance of victory without dynastic connections, but that more than doubles to 23% if the candidate does have such ties. By contrast, a KMT nominee who had already held, say, a township council post and ran for county council would have slightly better than even chances of winning even without a family legacy (55%). With a legacy on top of all that, she would be heavily favoured, with a 77% likelihood of success.

Figure 2. Predicted probability of winning

All other things being equal, the model assesses the dynastic advantage to be worth roughly a 20 percentage point boost in win probability, on average; the 95% confidence interval on this estimate ranges from 13 to 26%.²⁴ To put that in perspective, this is around the same magnitude as the advantage that major-party nominees enjoyed relative to others. Those running under the KMT banner also received about a 20 percentage point bump over those with neither KMT nor DPP nomination.²⁵ The dynastic advantage seems to be larger than that gained from having previously held elected office, which was worth an estimated 11 percentage points.²⁶ And it is around twice as large as the largest effect from any associational connection. Affiliation with a farmers or fishermans association netted candidates 10 percentage points, and local volunteering groups 6 percentage points.²⁷ As the confidence intervals make clear, these numbers should not be taken as precise certainties; instead they convey a sense of the likely magnitude of each effect.²⁸

All told, although several factors are correlated with a candidate's likelihood of winning a council election, the legacy advantage strongly holds its own in the regression analysis. As we have seen, dynastic candidates do not always win, but they enjoy much more favourable odds. Although many of them come to the race with deficits, such as being very young, lacking prior elected office or even less critical factors such as not having served in a grassroots volunteering capacity, the legacy advantage often helps them overcome these deficiencies.

5. Conclusion

It is well-established that candidates from political dynasties are more electorally successful than their counterparts from non-dynastic backgrounds. This is usually argued to be a result of advantages that dynastic candidates have in campaign-related resources, such as money, organizational networks and know-how, or brand-related advantages, such as voters' preference for familiar names, or both. However, the literature generally ignores variation in dynastic candidates' backgrounds. Some candidates come to the election with better-developed resumes than others. That is, some candidates have built up experiences, networks, expertise and connections with voters that can help them win elections. In this paper, we ask whether dynastic candidates' superior electoral performance is due to their training and experience or their political inheritance.

As expected, we find evidence that candidate quality generally matters. Candidates who are between the ages of 30 and 65, are major party nominees, have previously won an election and have held a position in a farmers or fisherpersons association or in a grassroots volunteer group all do better. We find evidence that candidate quality can explain some of the dynastic advantage. But the qualifications in which dynastic candidates are relatively strong are not ones that help a candidate win. Dynastics have a slight edge over others in terms of education, but this does not predict victory. Legacy candidates often serve as political aides and assistants, indeed they do so at much higher rates than non-legacy candidates, presumably by availing themselves of an inside track to work in their family member's office. But this does not correlate with success at the ballot box either.

Instead, legacy candidates enjoy special advantages that often cancel out or outweigh what would otherwise be weaknesses in their profile. They tend to be younger, and while non-legacy candidates in their 20s or early 30s are not very likely to win, legacy candidates still have an excellent shot at winning even in those age categories. Dynastics, moreover, tend to lack prior experience as elected officials, often jumping into city and county council elections without having bothered to win a lower office first. They also are somewhat less likely to have held posts in farmers or fisherpersons organizations, or to have volunteered at the grassroots level. Despite these weaknesses, they perform well because they enjoy compensating advantages.

Even after accounting for these differences in candidate quality, we find that dynastic status remains a powerful predictor of electoral success. Roughly speaking, the probability a dynastic candidate will win is about 20 percentage points higher than that for a similar non-dynastic candidate, though the exact number depends on the candidate's individual qualities. In short, while a well-developed resume might help, being a legacy candidate helps more.

Theoretically, this paper's focus on the diversity of legacy candidates requires a reconsideration of how and when electoral resources are transferred from one generation to another. We must bear in mind the fact that the category of ‘second-generation politicians’ includes a wide range of individuals. Previous literature assumes a brute force application of resources during the election campaign, with the established politician devising strategies, spending money and coordinating networks in order to win the election for the novice legacy candidate who brings little in terms of his or her own experiences or qualifications to the campaign. Yet dynastic candidates are not all alike, and the legacy inheritance takes different forms, conveyed at different times. Chiang Chao-kuo, for example, hardly had a seat in the Taichung City Council handed to him on a silver platter. He first ran 12 years after his father lost his seat, after having accumulated political experience in several positions at the national level, and he won more than twice the number of votes of his father's best campaigns. For some dynastic candidates, then, a portion of the inheritance takes place long before the election campaign. They may, even in childhood, acquire familiarity with the ways of politics: how supporters are cultivated, how a campaign headquarters is set up. In some cases, legacy candidates take it upon themselves to pursue experiences and connections that set up a successful plunge into electoral politics, rather than simply inheriting these things.

Yet Chiang is not necessarily typical, and in many other cases, the transfer of resources takes place only during and shortly before the campaign. Familial status opens doors that are not available to most other young people with ambitions for a future political career, and thus newcomers with such legacies are able to mount a campaign at a younger age than other candidates, and with a weaker resume – as in Lin's case. Dynastic candidates are able to do so well when entering politics so young precisely because they can rely on the campaign-specific inheritance to compensate for any remaining political deficiencies.

For Taiwan's local politics, the 2007 revision of the Local Government Act creating the new or enlarged Taichung, Tainan, New Taipei, Kaohsiung and, eventually, Taoyuan direct municipalities may result in more dynastic politicians. Candidate quality matters, and we find only two variables, a major party nomination and having previously held electoral office, on par with or close to dynastic status in impact. The 2007 reform eliminated township-level elections, the most important source of lower-level offices, for jurisdictions containing over 40% of the population. Without competition from town councillors and mayors, dynastic candidates for city and county councils face fewer obstacles (Batto, 2018a). The two major parties could change this dynamic by nominating more non-dynastic challengers, but they have not yet taken any steps to systematically reduce the number of legacy candidates winning party nominations.

This analysis raises interesting questions that we have not explored in this paper. One is whether the type of relationship with the dynastic predecessor matters. Are sons and daughters more successful than nieces and nephews? Are spouses more successful than brothers and sisters? Another involves whether dynastic candidates' experiences affect their careers in office and future electoral prospects. Recall our two running examples. After winning her race in 2014, Lin Ting-chun let it be known that she would not run for re-election. Under urging from the KMT, she eventually relented, but she lost the 2018 race.²⁹ Chiang Chao-kuo, in contrast, won his re-election race with the second most votes in the district, and cruised to a third term in 2022 as well, setting up a run for the national legislature in 2024.³⁰ Chiang's career seems to reflect things about the kind of legacy candidate he is: one who aimed for long-term success by gradually building qualifications that went well beyond whatever campaign resources his father handed off to him in 2014. Lin's shorter tenure, meanwhile, may relate to the relatively abbreviated process through which she was recruited, in which her uncle's well-meaning desire to endow a successor played an outsized role. More generally, while dynastic candidates with thinner resumes might be able to win office due to their political inheritance, it is not clear that that same inheritance equips them to perform well once in a council or when running for higher office – implying issues that can be examined in future research.